CN113434760A - Construction method recommendation method, device, equipment and storage medium - Google Patents

Construction method recommendation method, device, equipment and storage medium Download PDF

Info

Publication number
CN113434760A
CN113434760A CN202110711860.6A CN202110711860A CN113434760A CN 113434760 A CN113434760 A CN 113434760A CN 202110711860 A CN202110711860 A CN 202110711860A CN 113434760 A CN113434760 A CN 113434760A
Authority
CN
China
Prior art keywords
project
construction method
construction
text
engineering
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202110711860.6A
Other languages
Chinese (zh)
Other versions
CN113434760B (en
Inventor
杨婉琪
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An International Smart City Technology Co Ltd
Original Assignee
Ping An International Smart City Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An International Smart City Technology Co Ltd filed Critical Ping An International Smart City Technology Co Ltd
Priority to CN202110711860.6A priority Critical patent/CN113434760B/en
Publication of CN113434760A publication Critical patent/CN113434760A/en
Application granted granted Critical
Publication of CN113434760B publication Critical patent/CN113434760B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/90—Details of database functions independent of the retrieved data types
    • G06F16/95—Retrieval from the web
    • G06F16/953—Querying, e.g. by the use of web search engines
    • G06F16/9535—Search customisation based on user profiles and personalisation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F40/00—Handling natural language data
    • G06F40/30—Semantic analysis

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Computational Linguistics (AREA)
  • General Health & Medical Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Artificial Intelligence (AREA)
  • Data Mining & Analysis (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention relates to the field of artificial intelligence and discloses a construction method recommendation method, equipment and a storage medium. The method comprises the following steps: acquiring an engineering method content subset in an engineering method book by utilizing a natural language algorithm and an image segmentation technology, and performing applicability matching according to engineering name information and the engineering method content subset by analyzing a project file to be analyzed to obtain an optimal recommended engineering method of the engineering method book, wherein the engineering method corresponding to the engineering method content subset with the highest applicability is obtained; and matching construction integrity according to the project name information in the project file and the historical project information base information to obtain the historical project file with the highest construction integrity, wherein the construction method is the optimal recommended construction method for the historical project, and finally, the two construction methods are output. Based on the method, the recommended construction method does not need to read a large amount of data for comparison, the phenomenon of wrong comparison or missing of schemes in the process is avoided, the construction method recommendation efficiency is improved, and the matching accuracy is higher.

Description

Construction method recommendation method, device, equipment and storage medium
Technical Field
The invention relates to the field of artificial intelligence, in particular to a construction method recommendation method, a construction method recommendation device, construction equipment and a storage medium.
Background
With the development of scientific technology, especially for the technical application in the construction of engineering projects, artificial intelligence is gradually applied to the construction and analysis of engineering projects, such as NLP (natural language processing), the application of NLP in engineering is mainly the identification of files, but the identification application of NLP in engineering only consists in the simple identification and error correction of files, and for the artificial intelligence application of a construction method in the analysis of engineering files, no good implementation scheme exists at present.
At present, a construction scheme or construction method is recommended for a certain project, which is mainly implemented by reading and analyzing a bid document of the project or a project document which specifies the specific specification of the project construction manually by service personnel to obtain key contents, then, based on the key contents, related project information files in historical projects which have been under construction or completed construction are consulted, and a construction method or construction scheme which meets the construction requirements of the current bid project is screened from the historical projects, and in the process of consulting data, the service personnel needs to synthesize a large amount of historical project text information to repeatedly analyze to obtain a conclusion.
Disclosure of Invention
The invention mainly aims to solve the technical problem that the construction method recommendation efficiency and the matching rate are low due to the fact that the analysis process of the conventional engineering project construction method recommendation scheme is complex.
The invention provides a construction method recommendation method in a first aspect, which comprises the following steps: acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set; respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set; classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text; acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information; determining the corresponding relation between each project name in the project name set and project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the optimal recommended construction method of the construction method book of the project file text based on the applicability; according to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database; analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity; and outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
Optionally, in a first implementation manner of the first aspect of the present invention, the obtaining a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set includes: identifying a standard file in a book in an engineering method by using a file identification technology, calling a preset scanning program to scan the standard file, and outputting picture data of the standard file; carrying out structuring processing on the picture data by using an image segmentation model to obtain a structured file; identifying marking information in the structured file by using an image segmentation model, and cutting the picture data based on the marking information to obtain an engineering method table, an engineering method picture and an engineering method text; and combining the working method table, the working method picture and the working method text to obtain a text image block set.
Optionally, in a second implementation manner of the first aspect of the present invention, the performing a structuring process on the picture data by using an image segmentation model to obtain a structured file includes: identifying table content, image content and text content in the picture data by using an image segmentation model; recording the position information of the table content, the image content and the text content in the picture data; and labeling the table content, the image content and the text content according to the position information to obtain a structured file.
Optionally, in a third implementation manner of the first aspect of the present invention, the identifying and extracting the text image block set by using a natural language algorithm and a preset method name extraction model respectively to obtain the method name of each file in the text image block set includes: identifying the field information of the centralized construction method of the text image blocks by using a natural language algorithm; and extracting the working method words from the working method field information according to a preset working method name identification model to obtain the working method names of the files in the text image block set.
Optionally, in a fourth implementation manner of the first aspect of the present invention, the obtaining a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information includes: acquiring a project file to be analyzed from a management project system, and extracting text information in the project file to be analyzed by using a file identification technology to obtain a project file text; identifying the project name information in the project file text according to a preset project name identification model to obtain the project name information contained in the project file text; and constructing a project name set based on the project name information contained in the project file text, wherein the project name set contains at least one project name information.
Optionally, in a fifth implementation manner of the first aspect of the present invention, the determining, according to the project name set, a corresponding relationship between each project name in the project name set and project information in an engineering content subset, calculating a suitability between the project information in the engineering content subset and the project name information in the project name set, and determining an optimal recommended engineering method for an engineering book of the project file text based on the suitability: taking the project name of the project name set as a search term, and searching each engineering method content subset in the engineering method content database in a full-text search engine according to the search term to obtain a target project name, wherein the target project name is the project name in the engineering method content subset, and the target project name is the project name corresponding to the project information set; calculating the number of target engineering names in the content subsets of each construction method to obtain the applicability; and comparing the suitability degrees of the content subsets of the working methods, obtaining the content subset with the highest suitability degree based on the comparison result, and taking the working method corresponding to the content subset of the working method as the optimal recommended working method of the book of the working method.
Optionally, in a sixth implementation manner of the first aspect of the present invention, the optimal recommendation method for analyzing the building integrity of the project in the historical project data and determining the historical project of the project file text based on the building integrity by using a preset project building analysis model includes: according to a preset project construction analysis model, carrying out emotion word recognition on the historical project data to obtain construction emotion evaluation words of the historical project data; and carrying out emotion analysis on the constructed emotion evaluation words to obtain the strength of a positive word, and determining the construction integrity of the construction effect of each historical project according to the strength of the positive word, wherein the constructed emotion evaluation words comprise positive words and non-positive words. And comparing the construction integrity of the construction effect of each historical project, and taking the target construction method in the historical project as the optimal recommended construction method of the historical project, wherein the target construction method is the construction method in the historical project with the highest construction integrity.
The invention provides a construction method recommending device in a second aspect, which comprises the following steps: the segmentation module is used for acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set; the extraction module is used for respectively identifying and extracting the text image block set by utilizing a natural language algorithm and a preset construction method name extraction model to obtain the construction method name of each file in the text image block set; the classification module is used for classifying each file in the text image block set according to the construction method name to obtain a construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text; the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a project file text to be analyzed, extracting project name information in the project file text and generating a project name set based on the project name information; the calculation module is used for determining the corresponding relation between each project name in the project name set and the project information in the engineering content subset according to the project name set, calculating the applicability of the project information in the engineering content subset and the project name information in the project name set, and determining the optimal recommended engineering method of the engineering book of the project file text based on the applicability; the query module is used for querying historical project data matched with each project name information in the project name set from a historical project database according to the project name set; the integrity analysis module is used for analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity; and the recommending module is used for outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
Optionally, in a first implementation manner of the second aspect of the present invention, the dividing module includes: the identification unit is used for identifying a standard file in a book in an engineering method by using a file identification technology, calling a preset scanning program to scan the standard file and outputting picture data of the standard file; the image segmentation module is used for performing image segmentation on the image data to obtain image data; the cutting unit is used for identifying marking information in the structured file by using an image segmentation model and cutting the picture data based on the marking information to obtain an engineering method table, an engineering method picture and an engineering method text; and the merging unit is used for merging the construction method table, the construction method picture and the construction method text to obtain a text image block set.
Optionally, in a second implementation manner of the second aspect of the present invention, the structural processing unit is specifically configured to: identifying table content, image content and text content in the picture data by using an image segmentation model; recording the position information of the table content, the image content and the text content in the picture data; and labeling the table content, the image content and the text content according to the position information to obtain a structured file.
Optionally, in a third implementation manner of the second aspect of the present invention, the extraction module is specifically configured to: identifying the field information of the centralized construction method of the text image blocks by using a natural language algorithm; and extracting the working method words from the working method field information according to a preset working method name identification model to obtain the working method names of the files in the text image block set.
Optionally, in a fourth implementation manner of the second aspect of the present invention, the acquisition module is specifically configured to: acquiring a project file to be analyzed from a management project system, and extracting text information in the project file to be analyzed by using a file identification technology to obtain a project file text; identifying the project name information in the project file text according to a preset project name identification model to obtain the project name information contained in the project file text; and constructing a project name set based on the project name information contained in the project file text, wherein the project name set contains at least one project name information.
Optionally, in a fifth implementation manner of the second aspect of the present invention, the calculation module is specifically configured to: taking the project name of the project name set as a search term, and searching each engineering method content subset in the engineering method content database in a full-text search engine according to the search term to obtain a target project name, wherein the target project name is the project name in the engineering method content subset, and the target project name is the project name corresponding to the project information set; calculating the number of target engineering names in the content subsets of each construction method to obtain the applicability; and comparing the suitability degrees of the content subsets of the working methods, obtaining the content subset with the highest suitability degree based on the comparison result, and taking the working method corresponding to the content subset of the working method as the optimal recommended working method of the book of the working method.
Optionally, in a sixth implementation manner of the second aspect of the present invention, the complete analysis module is specifically configured to: according to a preset project construction analysis model, carrying out emotion word recognition on the historical project data to obtain construction emotion evaluation words of the historical project data; performing emotion analysis on the built emotion evaluation words to obtain the strength of positive words, and determining the building integrity of the construction effect of each historical project according to the strength of the positive words, wherein the built emotion evaluation words comprise positive words and non-positive words; and comparing the construction integrity of the construction effect of each historical project, and taking the target construction method in the historical project as the optimal recommended construction method of the historical project, wherein the target construction method is the construction method in the historical project with the highest construction integrity.
The third aspect of the present invention provides an engineering method recommendation apparatus, including: a memory and at least one processor, the memory having instructions stored therein; the at least one processor invokes the instructions in the memory to cause the construction method recommendation apparatus to perform the method of the construction method recommendation method described above.
A fourth aspect of the present invention provides a computer-readable storage medium having stored therein instructions, which when run on a computer, cause the computer to perform the steps of the above-described method of construction method recommendation.
According to the technical scheme provided by the invention, the image segmentation technology is utilized to segment the text of the books in the construction method, the contents in the graph are identified by the segmented text graph blocks through a natural language algorithm, and then the identified contents corresponding to different construction methods are segmented and classified to obtain the construction method content subsets corresponding to the names of the construction methods.
The method comprises the steps of obtaining a project file text to be analyzed, extracting project name information in the text to form a project name set, searching the applicability of the project name information between the project name information in the project name set and an engineering method content subset by a keyword searching method, taking the engineering method corresponding to the engineering method content subset with the highest applicability as the optimal recommended engineering method of an engineering method book, matching and corresponding the project information in the project file and an engineering method database, and selecting the engineering method with higher matching degree by a person with higher pertinence, so that the engineering method recommendation efficiency can be accelerated, and the engineering method matching recommendation rate of an engineering method book can be improved.
The method comprises the steps of utilizing project name information in project name concentration and historical project data in a historical project database to match the project name information to obtain historical project projects exceeding preset applicability values, utilizing a preset project construction analysis model to conduct construction effect analysis on the historical project exceeding the preset applicability values to obtain the historical project projects with the highest construction integrity, using an adopted construction method as an optimal recommendation construction method of the historical project, and finally outputting the obtained two construction methods.
Drawings
FIG. 1 is a schematic diagram of a first embodiment of a method for recommending an operating method according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a second embodiment of a method for recommending an engineering method according to an embodiment of the present invention;
FIG. 3 is a schematic diagram of a third embodiment of a method for recommending an operating method according to an embodiment of the present invention;
FIG. 4 is a schematic diagram of a fourth embodiment of a method for recommending an operating method according to an embodiment of the present invention;
FIG. 5 is a schematic diagram of a fifth embodiment of a method for recommending an operating method according to an embodiment of the present invention;
FIG. 6 is a schematic view of an embodiment of a construction method recommending device according to an embodiment of the present invention;
FIG. 7 is a schematic view of another embodiment of a construction method recommending apparatus according to an embodiment of the present invention;
FIG. 8 is a schematic view of an embodiment of a construction method recommending device in an embodiment of the present invention.
Detailed Description
The embodiment of the invention provides an engineering method recommending method, equipment and a storage medium. The method comprises the steps of segmenting text contents of a construction method book by a graph segmentation technology, then carrying out content identification on segmented text image blocks by using a natural language algorithm, carrying out segmentation and classification processing on identified different construction method contents again to obtain each construction method content subset, carrying out suitability degree matching on the engineering name information based on the obtained construction method content subset and a project file to be analyzed to obtain an optimal recommendation construction method of the construction method book, further carrying out suitability degree matching by using a historical project database and the project name information, and then carrying out construction integrity degree analysis by using a project construction analysis model to obtain the optimal recommendation construction method of the historical project.
The terms "first," "second," "third," "fourth," and the like in the description and in the claims, as well as in the drawings, if any, are used for distinguishing between similar elements and not necessarily for describing a particular sequential or chronological order. It will be appreciated that the data so used may be interchanged under appropriate circumstances such that the embodiments described herein may be practiced otherwise than as specifically illustrated or described herein. Furthermore, the terms "comprises," "comprising," or "having," and any variations thereof, are intended to cover non-exclusive inclusions, such that a process, method, system, article, or apparatus that comprises a list of steps or elements is not necessarily limited to those steps or elements expressly listed, but may include other steps or elements not expressly listed or inherent to such process, method, article, or apparatus.
For the convenience of understanding, the following describes a specific process of an embodiment of the present invention, and referring to fig. 1, a first embodiment of a method for recommending a construction method according to an embodiment of the present invention includes:
101. acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
in the step, the standard files may be construction principle books of an engineering method or construction specification files of the engineering method generated after the construction of an actual project, the files may be obtained from a Chinese construction project database system, that is, a background database in the Chinese construction project database system is accessed through a network interface to obtain the standard files (i.e., the construction specification files of the engineering method) corresponding to each construction method, then the engineering content in the standard files is identified and labeled by using an image identification technology or a text extraction technology, further, the labels are identified and labeled by using an image segmentation model, the image blocks are obtained by extracting according to the labeled content, the image segmentation model is a model trained by using an image contour identification algorithm, after all the labels in the standard files are completely extracted, and classifying the extracted image blocks to obtain a text image block set, wherein the text image block set comprises an engineering method table subset, an engineering method picture subset and an engineering method text subset.
In practical application, when classifying the extracted image blocks, the method specifically includes the following image blocks of several content types: the method comprises the steps of firstly utilizing an OpenCV function to conduct image gray processing, namely binarization processing, corroding and expanding to obtain a table line, then identifying and dividing each table in a standard file to obtain a table image block, further identifying and dividing the remaining content, obtaining image feature vectors and text feature vectors according to the standard file to serve as the basis of text types of predicted images, obtaining text image blocks in the standard file, identifying and processing the tables and the remaining division blocks after text identification into image blocks, and finally sorting the identified table image blocks, image blocks and text image blocks into a text image block set.
102. Respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set;
in this embodiment, based on the text image block set obtained in step S10, a natural language processing technique is used to identify table image blocks, text image blocks, and picture image blocks in the text image block set, and identify an engineering method corresponding to the content of the image blocks, where at least one of the image blocks includes an engineering method.
In the actual application process, the natural language technical algorithm analyzes and processes the text image blocks based on the natural language combining the stem knowledge base of the method and the logic inference rule, combines the inherent relation of the stem attribute and the logic inference rule to realize the corresponding method for identifying the text content in the text image block set, and labels the corresponding method label on the identified text image blocks
And the corresponding engineering method labels are used as the corresponding engineering methods of the image blocks by identifying the corresponding same engineering methods of the last and next table images or text images which are connected with the image of the image blocks through a natural language processing algorithm, so that the corresponding engineering methods of all the three types of image blocks including the table image blocks, the text image blocks and the image blocks in the text image block set are obtained.
The image blocks corresponding to the image blocks obtained through natural language algorithm processing are extracted according to the method name extraction model, the image blocks corresponding to two or more than two methods in the image blocks are extracted according to the method name extraction model, so that a single method corresponding to each image block is obtained, corresponding processing steps and method contents are provided based on each method, so that the overall method contents of each method are different, the text contents in the two or more than two image blocks are identified according to the method name identification model, the identified corresponding method contents are further cut to obtain the image blocks corresponding to the single method, and finally, each method table, method picture and method text are obtained.
103. Classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
in this embodiment, each of the working method tables, working method pictures, and working method texts in the text image block set obtained based on S20 are classified according to the working method name to obtain a working method content subset corresponding to each working method name, each working method image block is classified to call a classification processing function to classify the image block, the classification processing function firstly identifies and stores a number and a storage to the working method corresponding to each working method name image block based on a working method stem knowledge base, the same number is labeled to the same working method, the working method image blocks with the same number in the storage are traversed and classified into one class, all the same numbers in the storage are classified into one class by a function circulation mode, and the working method image blocks corresponding to the same number are further output together as one working method content subset to obtain a working method content subset corresponding to each working method name, wherein each working method content subset includes at least one working method table, and a working method table, At least one construction method picture and at least one construction method text.
104. Acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information;
in this embodiment, the project files are project files to be analyzed, which need to be matched with the corresponding working method for processing, and these files can be specifically obtained from a government affairs system of a current government bidding project, that is, a bidding project database in the government affairs system is accessed through a network interface to obtain project files needing to be matched with the corresponding working method, and then the project files to be analyzed are identified and converted into a project file text format (i.e., a normal Word document format) capable of being processed by the system by using a file identification technology.
Furthermore, based on the obtained project file text to be analyzed, the project name identification model is utilized to identify the project name in the project file text to be analyzed, the identified project name is collected into a project name set, the project name identification model extracts all the project names in the project file text to be analyzed by taking the feature information of the project name as the feature information of the extraction object, and then the extracted project names are collected into the project name set.
In practical application, because a project file text contains a plurality of characteristic information related to project targets, project requirements, project time limits, construction cost requirements and other project names, an existing project name database is called by a project name identification model to extract the project names in the project file text, the process firstly calls a project name and then edits the project file text, if the project file text is matched with the same project name, output is carried out, then a circulation function is called to circularly identify all the project names and traverse all the project names in the project file text, and finally all the project names identified by the project name identification model are collected into a project name set.
105. Determining the corresponding relation between each project name in the project name set and the project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the construction method recommendation construction method of the project file text based on the applicability;
in this embodiment, according to the project names included in the project name set obtained in S40, the corresponding relationship between each project name and the project information content in the construction method content subset is searched by using the keyword search method of the full text search engine using each project name as the keyword, the applicability of each construction method content subset is further obtained according to the quantity relationship between the keywords and the construction method content subsets, because each construction method content subset includes a lot of project name information, and some construction method content subsets even include the project name information of the feature information, based on this condition, each project name in the project name set is used as the keyword, and all the corresponding project names are searched in the construction method content subset by using the full text search engine.
In practical application, one project name in the project name set is used as a key word, the key word is searched in the content of all the construction method content subsets by using a key word searching method of a full-text search engine, the initial applicability of each construction method content subset is set to be 0, the applicability of the construction method content subset containing the key word obtained by searching is added by 1, further, each project name in the project name set is used as a key word, each key word is circularly traversed through all the construction method content subsets, and the applicability of the construction method content subset circularly traversed to the key word is added by 1 to obtain the applicability corresponding to each construction method content subset.
Setting a suitability threshold, obtaining the suitability of all the construction method content subsets after the circulation traversal is finished and comparing the suitability with the suitability threshold, obtaining the construction method content subsets exceeding the suitability threshold, further, comparing all the construction method content subsets exceeding the suitability threshold by using a comparison function to obtain the construction method content subset with the maximum suitability, the comparison function compares two of the suitability degrees of all the working method content subsets exceeding the suitability degree threshold value to obtain a working method content subset with a larger suitability degree, compares the working method content subset with the larger suitability degree with the suitability degree of the next working method content subset to obtain one of the larger suitability degrees, and finally obtaining an engineering method content subset with the highest applicability value by the cyclic comparison method, and taking the engineering method adopted by the engineering method content subset with the highest applicability value as the optimal recommended engineering method for the engineering method of the project file text.
106. According to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
in this embodiment, based on the project name set of the project file text to be analyzed obtained in S40, since each historical project includes the project characteristic information such as the target, the construction condition, the construction requirement, and the construction cost of the project, the project name in the project name set and the project information content in each historical project in the historical project database are used to perform matching correspondence of the project name information, so as to obtain the historical project data matched with each project name information in the project name set.
In practical application, one project name in the project name set of the project file text to be analyzed is taken as a keyword, the keyword is searched in the contents of all historical project projects by utilizing a keyword search method of a full-text search engine, setting the initial suitability of each historical project to be 0, adding 1 to the suitability of the historical project containing the keyword, further taking each project name in the project name set as the keyword, circularly traversing all the historical project by each keyword, adding 1 to the applicability of the historical project engineering which is circularly traversed to have the keyword to obtain the applicability corresponding to each historical project engineering, setting an applicability threshold, and outputting the historical project data with the historical project engineering suitability degree exceeding the suitability degree threshold after the loop traversal is finished.
107. Analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
in this embodiment, historical project engineering data exceeding an applicability threshold is obtained according to S60, an emotion analysis is performed on the construction effect evaluation in each historical project engineering in the historical project engineering data by using a historical project construction effect emotion analysis model, the construction integrity of the historical project engineering is obtained based on the strength of a positive term in the emotion analysis, the construction integrity values of the historical project engineering are compared, the historical project engineering with the highest construction integrity is obtained, and the construction method used by the historical project engineering is used as the optimal recommendation construction method for the historical project.
In practical application, according to the fact that the historical project engineering has the evaluation summary of the construction effect of the engineering, the evaluation summary of each project engineering in the historical project engineering data is subjected to emotion analysis by using a historical project construction effect emotion analysis model, and the processing method of the model comprises the following steps of: the construction effect evaluation of each historical project in the historical project engineering data is controlled to be a fixed length, and if the length is beyond the length cutoff, filling is carried out when the length is less than the length; the second stage is that the long and short term memory network is utilized to obtain the information of each comment, and the last output memory of the long and short term memory network is used as the semantic information of the whole sentence; and in the third stage, the vector is used as input and is sent into a classification layer for classification, so that the emotion analysis of the construction effect of the historical project is completed, and the strength of positive words in the construction effect obtained by the emotion analysis corresponds to the construction integrity of the corresponding historical project.
And comparing every two construction integrity values of each historical project in the historical project engineering database to obtain the historical project with the highest construction integrity, and further taking the construction method applied to the historical project as the optimal recommendation construction method of the historical project.
108. And outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
In this embodiment, two construction methods, that is, the optimal recommended construction method for the book as the construction method obtained in step S50 and the optimal recommended construction method for the history item obtained in step S70, are output to the engineer and reference construction selection is performed.
In the embodiment, the content of an engineering book text is segmented according to three forms of a text, an image and a table by using an image segmentation model, the obtained image blocks are used for identifying the content by using a natural language algorithm, the segmentation and classification processing are carried out on the content of different engineering methods obtained by identification, each engineering method content subset is obtained, the automatic acquisition of the engineering method in the engineering book can be realized, huge engineering method data are automatically obtained, the suitability matching of the engineering name information is carried out on the obtained engineering method content subsets and the project file to be analyzed, the optimal recommended engineering method of an engineering book is obtained, after the engineering method recommended book is obtained, the historical project engineering which actually obtains the application value is further obtained, the recommended engineering method has more practical significance value, after the suitability matching is carried out by using a historical project engineering database and the project name information, and analyzing the construction integrity by using the project construction analysis model to obtain the optimal recommended construction method of the historical project engineering. The method has the advantages that automatic matching and recommendation of the method books are achieved through the artificial intelligence technology, the method recommendation time can be greatly shortened, the matching degree of the method is higher, and two methods which meet requirements better are obtained.
Referring to fig. 2, a second embodiment of a method for recommending a construction method according to an embodiment of the present invention includes:
201. identifying a standard file in a book in an engineering method by using a file identification technology, calling a preset scanning program to scan the standard file, and outputting picture data of the standard file;
in this embodiment, the standard file refers to a standard normative Word document file of an engineering method book capable of being recognized and used by a system or an engineering method normative file generated by actual construction, that is, a large amount of engineering method data is obtained by accessing the engineering method data in a database system of a chinese construction project, the obtained engineering method data is processed by using a file recognition technology, the processing procedure is to uniformly convert different file format contents of the data in the engineering method data into a Word file format, abandon the file contents which cannot be recognized in the Word file format to obtain a standard file of the obtained engineering method data, and then scan and analyze the obtained standard file by using a common file scanning program, and the scanning program scans and outputs the standard file into a picture format in a text format to obtain picture data of the standard file.
202. Identifying table content, image content and text content in the picture data by using an image segmentation model;
in this embodiment, the obtained standard file includes three content formats, i.e., a table, an image and a text, so that the picture data obtained through the standard file is subjected to image segmentation model to identify the table content, the image content and the text content in the picture data, the table content is processed (i.e., binarized) by using an OpenCV function, the table line is identified to obtain a table image through the table line, the image feature vector and the text feature vector are further used as a text image prediction basis for the remaining picture data to obtain a text image block in the picture data, and the remaining picture data is identified as a picture image.
The model processing process comprises the steps of inputting acquired picture data into an image segmentation model to be trained when the image segmentation model is trained, obtaining the prediction probability of each type output by the image segmentation model, determining the weight coefficient of a preset loss function according to the number of the picture data, wherein the first weight coefficient comprises the ratio of the number of image sample data of a negative sample type to the number of image sample data of a positive sample type, namely a1＝(Ni/Nj)0.5(ii) a The second weight coefficient includes a ratio a of the number of image sample data of the positive sample class to the number of image sample data of the negative sample class2＝(Nj/Ni)0.5(where Ni is the number of image sample data in the negative sample class, and Nj is the number of image sample data in the positive sample class), correcting according to a preset loss function, i.e., -label × log (prediction) - α 2 × (1-label) × log (1-prediction) (where label is the labeling class of the artificially labeled training sample, and prediction is the prediction class output by the image segmentation model), determining a loss value, adjusting the model parameters of the image segmentation model according to the loss value, in which process the preset loss function includes a first evaluation part and a second evaluation part, and the first evaluation part determines a loss value of the prediction probability of the labeling class relative to 1; the second evaluation portion determines a loss value of the prediction probability of the non-labeled class with respect to 0. Through the first evaluation part and the second evaluation part, the loss value of the prediction probability of the labeled class relative to 1 and the loss value of the prediction probability of the non-labeled class relative to 0 can be respectively obtained, and the obtained loss values can be accurately obtainedReflecting the difference between the prediction probability and the expected probability, the accuracy of the classification of the trained model data can be improved.
203. Recording position information of table content, image content and text content in the picture data;
in this embodiment, in the identification process of the three image contents, position information is labeled on each identified image, the position information is labeled by adding image type numbers to page number diagonal rods of the identified page, where the image type numbers are table 1, image 2, and text 3, and then position information is labeled on all the table images, text images, and picture images obtained by identification.
204. Marking table content, image content and text content according to the position information to obtain a structured file;
in this embodiment, the obtained location information and the corresponding table content, image content, and text content are labeled to obtain a structured file.
205. Identifying annotation information in the structured file by using an image segmentation model, and cutting picture data based on the annotation information to obtain an engineering method table, an engineering method picture and an engineering method text;
in this embodiment, the graph partitioning model is used to cut three types of picture data of the structured file according to the standard position information thereof, all form images, picture images and text images are identified according to the image partitioning model, and corresponding contents are partitioned according to the position information thereof, so as to obtain each image block having position information.
206. Combining the working method table, the working method picture and the working method text to obtain a text image block set;
in this embodiment, all the obtained working method tables, working method pictures, and working method texts with position information are merged, and the same type is classified into a subset, so as to obtain a text image block set, which includes a working method table subset, a working method picture subset, and a working method text subset.
207. Respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set;
208. classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
209. acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information;
210. determining the corresponding relation between each project name in the project name set and the project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the optimal recommended construction method of the construction method book of the project file text based on the applicability;
211. according to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
212. analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
213. and outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
Wherein, the steps 207-213 are similar to the steps 102-108 described above, and detailed description thereof is omitted here.
In this embodiment, the detailed description is given to the image data of the standard file obtained by scanning, the table content, the image content, and the text content in the standard file are primarily cut by using the image segmentation model, three contents in the standard file are identified by using the image segmentation model, the relative positions of the identified contents are labeled, the labeled positions are cut, and all the tables, the images, and the texts obtained by cutting are combined to obtain the working method table subset, the working method text subset, and the working method image subset of the working method book. Compared with the prior art, the method can realize automatic identification and segmentation processing of a large amount of construction method data in the construction method book, greatly reduces identification and segmentation of the corresponding construction method when the construction method book data are recorded, and further improves the processing efficiency of the construction method data.
Referring to fig. 3, a third embodiment of a method for recommending a construction method according to an embodiment of the present invention includes:
301. acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
302. identifying field information of a method in the text image block set by using a natural language algorithm;
in this embodiment, a natural language algorithm is called for processing an engineering method table subset, an engineering method picture subset, and an engineering method text subset in an obtained text image block set, the algorithm belongs to the field of artificial intelligence, a model is firstly established for training the natural language algorithm, and the recognition accuracy of the algorithm is enhanced through training, and the training method includes: obtaining algorithm model training corpuses, extracting the characteristics of the training corpuses of natural language understanding tasks such as an engineering method stem knowledge base related to each engineering method, respectively carrying out model training on a preset multilayer perceptron model according to the characteristics of the training corpuses of the engineering method natural language understanding tasks to obtain a natural language understanding submodel corresponding to the engineering method natural language understanding tasks, and obtaining an engineering method natural language understanding model according to the natural language understanding submodel. Combining the trained natural language algorithm with the internal relation of the stem attribute and the logic separation rule, identifying text images in the content of the three image centralized construction methods to obtain construction methods in each text image block, labeling the identified construction methods according to the obtained position information labels, labeling the construction methods as the position information of the original image block, adding a diagonal bar and the construction method name of the corresponding content, further continuing to determine the construction method corresponding to each table image and perform standard position information on the table image based on the text information contained in the image and the construction method corresponding to the previous section of text corresponding to the text information, and determining the right-of-employment method for the rest images according to the corresponding previous section of text after the table processing is finished to obtain the construction method field information in the construction method table subset, the image subset and the construction method text subset.
303. Extracting construction method words from construction method field information according to a preset construction method name identification model to obtain construction method names of all files in a text image block set;
the method content in the method table subset, the method picture subset and the method text subset is identified by the method name identification model, the model obtains three types of subsets through a natural language algorithm, the method name identification model extracts the subsets with the same method name according to the position information of the subsets, content subsets included in the three types of subsets of the method are obtained, and the method names in the text image block subset are identified to obtain the method name of each file in the text image block subset.
304. Classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset into a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
305. acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information;
306. determining the corresponding relation between each project name in the project name set and the project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the optimal recommended construction method of the construction method book of the project file text based on the applicability;
307. according to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
308. analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
309. and outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
The steps 301 and 304 and 309 are similar to the steps 101 and 103 and 108, and detailed description thereof is omitted here.
This embodiment is based on the second embodiment; the method comprises the steps of utilizing a natural language algorithm to carry out recognition processing operation on contents in pictures of all obtained working method table subsets, working method text subsets and working method picture subsets, further realizing automatic recognition processing on the pictures, utilizing the natural language algorithm to carry out recognition training on working method word stems, further obtaining working method type natural language recognition models, enhancing recognition accuracy rate of the working method picture contents through continuous training of the models, further accurately recognizing contents corresponding to different working methods in the picture contents by utilizing the algorithm, and carrying out labeling processing in the recognized corresponding working method contents, thereby obtaining the working method tables, the working method pictures and the working method texts included by each working method. Compared with the prior art, the method can automatically identify the content of the divided large number of the pictures of the construction method, lays a foundation for classifying the corresponding content by different construction methods, greatly accelerates the classification processing of different construction methods in the books of the construction method, and realizes the automatic identification of the pictures of the construction method.
Referring to fig. 4, a fourth embodiment of the method for recommending a construction method according to the embodiment of the present invention includes:
401. acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
402. respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set;
403. classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset into a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
404. acquiring a project file to be analyzed from a management project system, and extracting text information in the project file to be analyzed by using a file identification technology to obtain a project file text;
in this embodiment, the project file to be analyzed is a bidding project file to be subject to construction method matching, the file is obtained from a government website and government affairs system, that is, the government affairs project system is accessed through a network interface to obtain files of projects to be bid, then the files are identified and converted into a Word document processing format of the system by using a file identification technology, in the identification and conversion process, the text in the files is accurately identified and converted, the form in the files accurately identifies form lines and contents in the forms, and the pictures in the forms are converted without distortion to obtain the project file text of the files to be analyzed, wherein the text and the forms are accurate and the pictures are clear.
405. Identifying the project name information in the project file text according to a preset project name identification model to obtain the project name information contained in the project file text;
in the embodiment, the project names in the text content of the project files are identified by using a project name identification model, each bidding project file contains project name, project specification, cost requirement and other project name characteristic information, the project file can be known and constructed by extracting the project name information, the project name identification model stores and trains a large amount of project name information through an existing project name database, the project name information in the project files is identified, and the identification is realized by calling the project name information in the model database and traversing all the project name information by using a circulation function so as to obtain all the project name information contained in the project file text.
406. Constructing a project name set based on project name information contained in a project file text, wherein the project name set contains at least one project name information;
in this embodiment, all the obtained project name information is stored in a set, and a project name set is constructed, where the project name set at least includes one project name information.
407. Taking the project name of the project name set as a search term, and searching each engineering method content subset in an engineering method content database in a full-text search engine according to the search term to obtain a target project name, wherein the target project name is the project name in the engineering method content subset, and the target project name is the project name corresponding to the project information set;
in this embodiment, the keyword is each project name in the project name set, that is, each project name in the obtained project name set is used as a search keyword, and the search method of keyword search of the full-text search engine is used to search the project names in all the working method content subsets.
In the specific operation steps, each project name of the project name set obtained above is used as a search key word, a search method of key word search of a full-text search engine is utilized, when a key word is input into a system for query, the search engine can find indexes of all related construction method contents which accord with the key word from a huge construction method content subset database, the key word is subjected to full-text traversal search in the established construction method content subset database to obtain a corresponding relation containing the key word in the construction method content subset, one project name information is taken as the key word to be searched in each construction method in all the construction method content subsets through a function circulation method for corresponding, all the construction method content subsets are retrieved in a circulation traversal mode, and the corresponding relation between the construction method content subsets and each project name in the project name set is output.
408. Calculating the number of target engineering names in the content subsets of each construction method to obtain the applicability;
in this embodiment, when the database of working method content subsets traverses all working method content subsets by using a keyword search method, setting an initial value of the applicability of each working method content subset to be 0 (the applicability is recorded as the number of keywords contained in the working method content subset), retrieving the database of the working method content subsets to obtain the working method content subsets with the keywords, searching the applicability of the working method content subsets with the keywords and adding 1 to process, traversing all the working method name keywords through a loop function to the database, and obtaining the applicability of each working method content subset.
409. Comparing the suitability degrees of the content subsets of the working methods, obtaining a working method content subset with the highest suitability degree based on the comparison result, and taking the working method corresponding to the working method content subset as the optimal recommended working method of the working method book;
in this embodiment, a comparison function is called according to the obtained applicability of each subset of the content subsets of the construction method to obtain a content subset of the construction method with the highest applicability, the step is to compare the content subsets of the construction method with the applicability higher than a threshold value in pairs by calling the comparison function, compare the content subset of the construction method with the high applicability with the content subset of the next construction method, and obtain the content subset of the construction method with the highest applicability until the comparison with the applicability of the content subset of the last construction method is finished, so that the construction method applied by the content subset of the construction method with the highest applicability is used as an optimal recommended construction method for construction method books.
410. According to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
411. analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
412. and outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
Wherein, the steps 401-.
In the embodiment, a project file text is obtained from a government affair system, all project name information in the project file text is identified through a project name identification model, all identified project name information forms a project name set, all construction method content subsets established are searched by using project names in the project name set as search keywords, and a construction method corresponding to the construction method content subset with the highest applicability is selected as an optimal recommendation construction method of a construction method book according to the number of the project name keywords contained in the construction method content subset searched as the applicability. Compared with the prior art, the method and the system have the advantages that the engineering information contained in the project file text is skillfully utilized, the keyword search of full-text retrieval is utilized, the number of the keywords is utilized to construct the applicability, so that the optimal recommended construction method is obtained, the construction method with high engineering matching degree can be obtained, and the search time can be greatly shortened to realize the search of a plurality of construction methods in the construction method books.
Referring to fig. 5, a fifth embodiment of a method for recommending a construction method according to an embodiment of the present invention includes:
501. acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
502. respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set;
503. classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
504. acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information;
505. determining the corresponding relation between each project name in the project name set and the project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the optimal recommended construction method of the construction method book of the project file text based on the applicability;
506. according to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
507. according to a preset project construction analysis model, emotion word recognition is carried out on historical project data to obtain construction emotion evaluation words of the historical project data;
calling a project construction effect sentiment analysis noun model algorithm according to the obtained target historical project data, carrying out sentiment word analysis on the construction effect in the target historical project, firstly preprocessing the data by the project construction effect sentiment analysis noun model, controlling the evaluation of the historical project engineering construction effect to be a fixed length for subsequent sentiment word recognition, controlling the fixed length of the evaluation of the construction effect to be 200 characters, cutting the evaluation exceeding the changed length into a plurality of sections for processing, supplementing the length to not more than the length in a space supplementing mode to 200 characters, secondly acquiring the information of each comment by using Lstm (namely a long-short time memory network), memorizing the last output of the long-short time memory network as the semantic information of the whole sentence, finally inputting the obtained statement information as a target vector, and classifying the sentiment words by sending the target vector into a classification layer, and the construction effect emotional words are deeply learned by Lstm to continuously optimize the parameters identified by the model, so that the emotional analysis of the construction effect of the historical project is completed.
508. Carrying out emotion analysis on the constructed emotion evaluation words to obtain the strength of the positive words, and determining the construction integrity of the construction effect of each historical project according to the strength of the positive words, wherein the constructed emotion evaluation words comprise positive words and non-positive words;
and determining the construction integrity of the historical engineering project based on the construction integrity of the historical engineering project obtained by the positive evaluation words, namely the positive strength of the positive words and the number of the positive words according to the emotion words obtained by classification analysis, wherein the emotion words comprise positive evaluation words and negative evaluation words.
509. And comparing the construction integrity of the construction effect of each historical project, and taking the target construction method in the historical project as the optimal recommended construction method of the historical project. The target construction method is a construction method for constructing a historical project with the highest integrity;
and comparing the building integrity of each historical engineering project in the obtained historical project data by calling a comparison function so as to obtain the historical engineering project with the highest building integrity, wherein the called comparison function is to compare the building integrity values of two historical engineering projects in the historical project engineering data to obtain the historical engineering project with the highest building integrity, and then compare the historical engineering project with the next historical engineering project until the historical engineering projects in all the historical project data are compared, so as to obtain the historical engineering project with the highest building integrity, and the construction method applied by the historical engineering project is used as the optimal recommendation construction method for the historical engineering project.
510. And outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
The steps 501, 506 and 510 are similar to the steps 101, 105 and 108, and are not described herein again.
In this embodiment, on the basis of the fourth embodiment, all the project name information of the project name set of the obtained project file text is used as search keywords, a full-text search method is used to perform full-text search on a historical project database, a historical project beyond a preset applicability is selected according to the number of the project name keywords contained in the retrieved historical project as the applicability, a project construction effect emotion analysis noun model algorithm is called, emotion word analysis is performed on the construction effect in the target historical project, the historical project with the highest construction integrity is obtained through analysis, and the construction method is used as the optimal recommendation construction method for the historical project. Compared with the prior art, on the basis of obtaining the recommended construction method of the construction method book, the historical project engineering with more practical value is further obtained, the method has more reference significance by combining the practical engineering method and the recommended construction method of the construction method book, the construction integrity of the historical project engineering is analyzed by considering the relevance of the engineering and applying an emotion analysis model in the selection process, and the selected historical project engineering has more practical reference value.
In the above description of the method for recommending an engineering method in an embodiment of the present invention, referring to fig. 6, a method for recommending an engineering method in an embodiment of the present invention is described below, where a first embodiment of the method for recommending an engineering method in an embodiment of the present invention includes:
the segmentation module 601 is configured to obtain a standard file in a book in an engineering method, and segment each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
an extraction module 602, which is used for respectively identifying and extracting the text image block set by using a natural language algorithm and a preset method name extraction model to obtain the method name of each file in the text image block set;
a classification module 603, configured to classify each file in the text image block set according to an engineering method name, obtain an engineering method content subset corresponding to each engineering method name, and store the engineering method content subset in an engineering method content database, where the engineering method content subset includes at least one engineering method table, at least one engineering method picture, and at least one engineering method text;
the acquisition module 604 is configured to acquire a project file text to be analyzed, extract project name information in the project file text, and generate a project name set based on the project name information;
a calculating module 605, configured to determine, according to the project name set, a corresponding relationship between each project name in the project name set and the project information in the engineering content subset, calculate a suitability between the project information in the engineering content subset and the project name information in the project name set, and determine an optimal recommended engineering method of an engineering book of the project file text based on the suitability;
the query module 606 is used for querying historical project data matched with each project name information in the project name set from the historical project database according to the project name set;
the integrity analysis module 607 is used for analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
and the recommending module 608 is used for outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
In the embodiment, the method recommending device for the method recommending device is operated, the method recommending device utilizes an image segmentation model to segment the content of a text of a method book according to three forms of the text, the image and the table, the obtained image blocks utilize a natural language algorithm to identify the content, the identified content of different methods is segmented and classified to obtain each method content subset, the method recommending device can automatically obtain the method in the method book, automatically obtain huge method data, perform the suitability matching of project name information based on the obtained method content subset and a project file to be analyzed to obtain the optimal recommended method of the method book, further obtain the historical project engineering actually obtained application value after obtaining the method recommending method of the book, and ensure that the recommended method has more practical significance, and after the suitability degree matching is carried out by utilizing the historical project engineering database and the project name information, the construction integrity degree is analyzed by utilizing a project construction analysis model to obtain the optimal recommended construction method of the historical project engineering. The method has the advantages that automatic matching and recommendation of the method books are achieved through the artificial intelligence technology, the method recommendation time can be greatly shortened, the matching degree of the method is higher, and two methods which meet requirements better are obtained.
Referring to fig. 7, a second embodiment of the construction method recommending apparatus according to the embodiment of the present invention specifically includes:
the segmentation module 601 is configured to obtain a standard file in a book in an engineering method, and segment each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
an extraction module 602, configured to respectively identify and extract the text image block set by using a natural language algorithm and a preset method name extraction model, so as to obtain a method name of each file in the text image block set;
a classification module 603, configured to classify each file in the text image block set according to an engineering method name, obtain an engineering method content subset corresponding to each engineering method name, and store the engineering method content subset in an engineering method content database, where the engineering method content subset includes at least one engineering method table, at least one engineering method picture, and at least one engineering method text;
the acquisition module 604 is configured to acquire a project file text to be analyzed, extract project name information in the project file text, and generate a project name set based on the project name information;
a calculating module 605, configured to determine, according to the project name set, a corresponding relationship between each project name in the project name set and the project information in the engineering content subset, calculate a suitability between the project information in the engineering content subset and the project name information in the project name set, and determine an optimal recommended engineering method of an engineering book of the project file text based on the suitability;
the query module 606 is used for querying historical project data matched with each project name information in the project name set from the historical project database according to the project name set;
the integrity analysis module 607 is used for analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
and the recommending module 608 is used for outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
In this embodiment, the segmentation module 601 may specifically include the following units:
the identification unit 6011 is configured to identify a standard file in a book by using a file identification technology, call a preset scanning program to scan the standard file, and output picture data of the standard file;
a structured processing unit 6012, configured to perform structured processing on the picture data by using the image segmentation model to obtain a structured file;
a clipping unit 6013, configured to recognize labeling information in the structured file by using the image segmentation model, and clip the picture data based on the labeling information to obtain an engineering method table, an engineering method picture, and an engineering method text;
and a merging unit 6014, configured to merge the working method table, the working method picture, and the working method text to obtain a text image block set.
The structured processing unit 6012 is specifically configured to:
identifying table content, image content and text content in the picture data by using an image segmentation model; recording position information of table content, image content and text content in the picture data; and marking the table content, the image content and the text content according to the position information to obtain a structured file.
The acquisition module 604 is specifically configured to: acquiring a project file to be analyzed from a management project system, and extracting text information in the project file to be analyzed by using a file identification technology to obtain a project file text; identifying the project name information in the project file text according to a preset project name identification model to obtain the project name information contained in the project file text; and constructing a project name set based on the project name information contained in the project file text, wherein the project name set contains at least one project name information.
The calculating module 605 is specifically configured to: taking the project name of the project name set as a search term, and searching each engineering method content subset in an engineering method content database in a full-text search engine according to the search term to obtain a target project name, wherein the target project name is the project name in the engineering method content subset, and the target project name is the project name corresponding to the project information set; calculating the number of target engineering names in the content subsets of each construction method to obtain the applicability; and comparing the suitability degrees of the content subsets of the working methods, obtaining the content subset of the working method with the highest suitability degree based on the comparison result, and taking the working method corresponding to the content subset of the working method as the optimal recommended working method of the book of the working method.
Wherein, the integrity analysis module 607 is specifically configured to: according to a preset project construction analysis model, emotion word recognition is carried out on historical project data to obtain construction emotion evaluation words of the historical project data; carrying out emotion analysis on the constructed emotion evaluation words to obtain the strength of the positive words, and determining the construction integrity of the construction effect of each historical project according to the strength of the positive words, wherein the constructed emotion evaluation words comprise positive words and non-positive words; and comparing the construction integrity of the construction effect of each historical project, and taking the target construction method in the historical project as the optimal recommended construction method of the historical project, wherein the target construction method is the construction method in the historical project with the highest construction integrity.
In the embodiment of the invention, the automatic recommendation of the construction method is realized by the construction method recommendation device, the image data of the input book construction method text is firstly segmented and identified by the image segmentation technology and the natural language technology, the content corresponding to more construction method content subsets is segmented and classified to form a construction method content subset database, and huge construction method book data is arranged into the construction method content subset database to facilitate the query of the subsequent project file construction method.
Furthermore, the optimal recommended construction method of the construction method book and the optimal recommended construction method of the historical project are obtained by identifying the matching of the adaptation degree and the construction integrity of the project name information in the project file text to be analyzed and the construction method content subset database and the historical project database, the project name information in the project file to be analyzed is fully utilized, the engineering information is utilized for matching, the construction method with higher related matching degree can be matched, in addition, the use experience of some construction methods can be provided for inquirers by combining the historical construction effect, and the practical reference value is higher.
Fig. 6 and 7 describe the construction method recommending device in the embodiment of the present invention in detail from the perspective of the modular functional entity, and the construction method recommending device in the embodiment of the present invention is described in detail from the perspective of hardware processing.
The following describes the construction method recommendation device in the embodiment of the invention in detail from the perspective of hardware processing.
Fig. 8 is a schematic structural diagram of a construction method recommendation device 800 according to an embodiment of the present invention, which may have a relatively large difference due to different configurations or performances, and may include one or more processors (CPUs) 810 (e.g., one or more processors) and a memory 820, and one or more storage media 830 (e.g., one or more mass storage devices) storing an application 833 or data 832. Memory 820 and storage medium 830 may be, among other things, transient or persistent storage. The program stored in the storage medium 830 may include one or more modules (not shown), each of which may include a series of instruction operations in the construction method recommending apparatus 800. Further, the processor 810 may be configured to communicate with the storage medium 830, and execute a series of instruction operations in the storage medium 830 on the construction method recommending apparatus 800.
The method recommendation apparatus 800 may also include one or more power supplies 840, one or more wired or wireless network interfaces 850, one or more input-output interfaces 860, and/or one or more operating systems 831, such as Windows Server, Mac OS X, Unix, Linux, FreeBSD, etc. Those skilled in the art will appreciate that the construction method recommended equipment configuration shown in FIG. 8 does not constitute a limitation of the construction method recommended equipment and may include more or fewer components than shown, or some components in combination, or a different arrangement of components.
The invention also provides an engineering method recommending device, which comprises a memory and a processor, wherein the memory is stored with computer readable instructions, and the computer readable instructions when executed by the processor cause the processor to execute the steps of the engineering method recommending method in the embodiments.
The present invention also provides a computer-readable storage medium, which may be a non-volatile computer-readable storage medium, and which may also be a volatile computer-readable storage medium, having stored therein instructions, which, when executed on a computer, cause the computer to perform the steps of the method for worker method recommendation.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described systems and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied in the form of a software product, which is stored in a storage medium and includes instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a read-only memory (ROM), a Random Access Memory (RAM), a magnetic disk, or an optical disk.
The above-mentioned embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the same; although the present invention has been described in detail with reference to the foregoing embodiments, it will be understood by those of ordinary skill in the art that: the technical solutions described in the foregoing embodiments may still be modified, or some technical features may be equivalently replaced; and such modifications or substitutions do not depart from the spirit and scope of the corresponding technical solutions of the embodiments of the present invention.

Claims (10)

1. A construction method recommendation method is characterized by comprising the following steps:
acquiring a standard file in a book in an engineering method, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set;
respectively identifying and extracting the text image block set by using a natural language algorithm and a preset construction method name extraction model to obtain construction method names of all files in the text image block set;
classifying each file in the text image block set according to the construction method name to obtain an construction method content subset corresponding to each construction method name, and storing the construction method content subset to a construction method content database, wherein the construction method content subset comprises at least one construction method table, at least one construction method picture and at least one construction method text;
acquiring a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information;
determining the corresponding relation between each project name in the project name set and project information in the construction method content subset according to the project name set, calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and determining the optimal recommended construction method of the construction method book of the project file text based on the applicability;
according to the project name set, historical project data matched with each project name information in the project name set are inquired from a historical project database;
analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
and outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
2. The method for recommending an engineering method according to claim 1, wherein the obtaining a standard file in an engineering book, and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set comprises:
identifying a standard file in a book in an engineering method by using a file identification technology, calling a preset scanning program to scan the standard file, and outputting picture data of the standard file;
carrying out structuring processing on the picture data by using an image segmentation model to obtain a structured file;
identifying marking information in the structured file by using an image segmentation model, and cutting the picture data based on the marking information to obtain an engineering method table, an engineering method picture and an engineering method text;
and combining the working method table, the working method picture and the working method text to obtain a text image block set.
3. The method for recommending engineering work according to claim 2, wherein said using an image segmentation model to perform a structuring process on said picture data to obtain a structured file comprises:
identifying table content, image content and text content in the picture data by using an image segmentation model;
recording the position information of the table content, the image content and the text content in the picture data;
and labeling the table content, the image content and the text content according to the position information to obtain a structured file.
4. The method for recommending an engineering method according to any one of claims 1-3, wherein the identifying and extracting the text tile set by using a natural language algorithm and a preset engineering method name extracting model respectively to obtain the engineering method name of each file in the text tile set comprises:
identifying the field information of the centralized construction method of the text image blocks by using a natural language algorithm;
and extracting the working method words from the working method field information according to a preset working method name identification model to obtain the working method names of the files in the text image block set.
5. The method for recommending construction methods according to claim 4, wherein the obtaining a project file text to be analyzed, extracting project name information in the project file text, and generating a project name set based on the project name information comprises:
acquiring a project file to be analyzed from a management project system, and extracting text information in the project file to be analyzed by using a file identification technology to obtain a project file text;
identifying the project name information in the project file text according to a preset project name identification model to obtain the project name information contained in the project file text;
and constructing a project name set based on the project name information contained in the project file text, wherein the project name set contains at least one project name information.
6. The method for recommending construction methods according to claim 4, wherein the determining a corresponding relationship between each project name in a project name set and project information in a construction method content subset according to the project name set, and calculating the applicability of the project information in the construction method content subset and the project name information in the project name set, and the determining the optimal recommendation construction method of the construction method book of the project file text based on the applicability comprises:
taking the project names of the project name set as search terms, and searching each engineering content subset in the engineering content database in a full-text search engine according to the search terms to obtain target project names, wherein the target project names are project names in the engineering content subsets, and the target project names are project names corresponding to the project information set;
calculating the number of target engineering names in the content subsets of each construction method to obtain the applicability;
and comparing the suitability degrees of the content subsets of the working methods, obtaining the content subset with the highest suitability degree based on the comparison result, and taking the working method corresponding to the content subset of the working method as the optimal recommended working method of the book of the working method.
7. The construction method recommendation method according to claim 4, wherein the analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model, and determining the optimal recommended construction method for the historical project of the project file text based on the construction integrity comprises:
according to a preset project construction analysis model, carrying out emotion word recognition on the historical project data to obtain construction emotion evaluation words of the historical project data;
performing emotion analysis on the built emotion evaluation words to obtain the strength of positive words, and determining the building integrity of the construction effect of each historical project according to the strength of the positive words, wherein the built emotion evaluation words comprise positive words and non-positive words;
and comparing the construction integrity of the construction effect of each historical project, and taking the target construction method in the historical project as the optimal recommended construction method of the historical project, wherein the target construction method is the construction method in the historical project with the highest construction integrity.
8. An engineering method recommending device, characterized in that the engineering method recommending device comprises:
the system comprises a segmentation module, a storage module and a processing module, wherein the segmentation module is used for acquiring a standard file in an engineering method book and segmenting each page of content in the standard file by using a preset image segmentation model to obtain a text image block set, and the text image block set comprises an engineering method table subset, an engineering method picture subset and an engineering method text subset;
the extraction module is used for respectively identifying and extracting the construction method names in each construction method table, construction method picture and construction method text in the construction method table subset, the construction method picture subset and the construction method text subset by utilizing a natural language algorithm and a preset construction method name extraction model to obtain the construction method names corresponding to each construction method table, construction method picture and construction method text;
the classification module is used for classifying the construction method table subset, the construction method picture subset and the construction method text subset according to the construction method names to obtain construction method content subsets corresponding to each construction method name, and storing the construction method content subsets into a construction method content database, wherein the construction method content subsets comprise at least one construction method table, at least one construction method picture and at least one construction method text;
the system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring a project file text to be analyzed, extracting project name information in the project file text and generating a project name set based on the project name information;
the calculation module is used for determining the corresponding relation between each project name in the project name set and the project information in the engineering content subset according to the project name set, calculating the applicability of the project information in the engineering content subset and the project name information in the project name set, and determining the optimal recommended engineering method of the engineering book of the project file text based on the applicability;
the query module is used for querying historical project data matched with each project name information in the project name set from a historical project database according to the project name set;
the integrity analysis module is used for analyzing the construction integrity of the project in the historical project data by using a preset project construction analysis model and determining the optimal recommended construction method of the historical project of the project file text based on the construction integrity;
and the recommending module is used for outputting the recommended construction method matched with the construction method book and the optimal recommended construction method of the historical project as the construction method of the project file to be analyzed.
9. An engineering method recommending device is characterized by comprising the following steps: a memory and at least one processor, the memory having instructions stored therein;
the at least one processor invokes the instructions in the memory to cause the construction method recommendation apparatus to perform the steps of the construction method recommendation method of any of claims 1-7.
10. A computer readable storage medium having instructions stored thereon, wherein the instructions, when executed by a processor, implement the steps of the method for worker recommendation as claimed in any one of claims 1 to 7.
CN202110711860.6A 2021-06-25 2021-06-25 Construction method recommendation method, device, equipment and storage medium Active CN113434760B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110711860.6A CN113434760B (en) 2021-06-25 2021-06-25 Construction method recommendation method, device, equipment and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110711860.6A CN113434760B (en) 2021-06-25 2021-06-25 Construction method recommendation method, device, equipment and storage medium

Publications (2)

Publication Number Publication Date
CN113434760A true CN113434760A (en) 2021-09-24
CN113434760B CN113434760B (en) 2022-12-27

Family

ID=77754519

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110711860.6A Active CN113434760B (en) 2021-06-25 2021-06-25 Construction method recommendation method, device, equipment and storage medium

Country Status (1)

Country Link
CN (1) CN113434760B (en)

Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108764610A (en) * 2018-04-12 2018-11-06 福建工程学院 Standard technology construction management of project method based on big data, storage medium
CN109614523A (en) * 2018-11-27 2019-04-12 建经投资咨询有限公司 A kind of engineering technology counseling services management system and method
CN109670783A (en) * 2018-12-05 2019-04-23 中铁四局集团第五工程有限公司 BIM-based construction enterprise intelligent library establishment method
JP2019066920A (en) * 2017-09-28 2019-04-25 オリンパス株式会社 Construction information acquirement apparatus, construction information management system, and information terminal
CN109685576A (en) * 2018-12-25 2019-04-26 华联世纪工程咨询股份有限公司 A kind of automatic coordination valence method, apparatus of engineering material, terminal device and storage medium
CN110619187A (en) * 2019-09-25 2019-12-27 三峡大学 Concrete dam construction scheme and resource conflict detection and adjustment method
CN111563173A (en) * 2020-05-09 2020-08-21 北京中科院软件中心有限公司 Knowledge graph construction method and system based on manufacturing field
CN112241623A (en) * 2020-09-17 2021-01-19 华中科技大学 Automatic generation device and method for contact network construction technology document content
CN112633857A (en) * 2021-01-04 2021-04-09 广联达科技股份有限公司 Construction progress recommendation method and device, computer equipment and readable storage medium

Patent Citations (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2019066920A (en) * 2017-09-28 2019-04-25 オリンパス株式会社 Construction information acquirement apparatus, construction information management system, and information terminal
CN108764610A (en) * 2018-04-12 2018-11-06 福建工程学院 Standard technology construction management of project method based on big data, storage medium
CN109614523A (en) * 2018-11-27 2019-04-12 建经投资咨询有限公司 A kind of engineering technology counseling services management system and method
CN109670783A (en) * 2018-12-05 2019-04-23 中铁四局集团第五工程有限公司 BIM-based construction enterprise intelligent library establishment method
CN109685576A (en) * 2018-12-25 2019-04-26 华联世纪工程咨询股份有限公司 A kind of automatic coordination valence method, apparatus of engineering material, terminal device and storage medium
CN110619187A (en) * 2019-09-25 2019-12-27 三峡大学 Concrete dam construction scheme and resource conflict detection and adjustment method
CN111563173A (en) * 2020-05-09 2020-08-21 北京中科院软件中心有限公司 Knowledge graph construction method and system based on manufacturing field
CN112241623A (en) * 2020-09-17 2021-01-19 华中科技大学 Automatic generation device and method for contact network construction technology document content
CN112633857A (en) * 2021-01-04 2021-04-09 广联达科技股份有限公司 Construction progress recommendation method and device, computer equipment and readable storage medium

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
周玉芳: "搜索引擎在工法科技查新中的应用", 《农业图书情报学刊》 *
李东风等: "基于BIM的水电工程工法管理***应用研究", 《中国建设信息化》 *
杨爱英等: "工法查新与一般科技查新的比较研究", 《图书情报工作》 *
杨铭等: "面向对象的施工企业工法库设计与建设研究", 《安徽建筑》 *

Also Published As

Publication number Publication date
CN113434760B (en) 2022-12-27

Similar Documents

Publication Publication Date Title
CN111222305B (en) Information structuring method and device
CN108932294B (en) Resume data processing method, device, equipment and storage medium based on index
US20220237230A1 (en) System and method for automated file reporting
CN112035599B (en) Query method and device based on vertical search, computer equipment and storage medium
CN113961685A (en) Information extraction method and device
CN115470338B (en) Multi-scenario intelligent question answering method and system based on multi-path recall
CN116628173B (en) Intelligent customer service information generation system and method based on keyword extraction
Van Phan et al. A nom historical document recognition system for digital archiving
CN115630843A (en) Contract clause automatic checking method and system
CN114491034B (en) Text classification method and intelligent device
CN111475607A (en) Web data clustering method based on Mashup service function characteristic representation and density peak detection
CN117592470A (en) Low-cost gazette data extraction method driven by large language model
CN116127105B (en) Data collection method and device for big data platform
CN117171650A (en) Document data processing method, system and medium based on web crawler technology
TW202207109A (en) Document management method and system for engineering project
CN111460114A (en) Retrieval method, device, equipment and computer readable storage medium
CN111104422A (en) Training method, device, equipment and storage medium of data recommendation model
CN113434760B (en) Construction method recommendation method, device, equipment and storage medium
CN114780712A (en) Quality evaluation-based news topic generation method and device
CN112613290A (en) Document template generation method, device, equipment and storage medium
CN111209375A (en) Universal clause and document matching method
CN115146630B (en) Word segmentation method, device, equipment and storage medium based on professional domain knowledge
CN116992874B (en) Text quotation auditing and tracing method, system, device and storage medium
CN113660322B (en) Offline cloud-sharing method and system
CN116681042B (en) Content summary generation method, system and medium based on keyword extraction

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant