CN110795567A - Knowledge graph platform - Google Patents
Knowledge graph platform Download PDFInfo
- Publication number
- CN110795567A CN110795567A CN201910933721.0A CN201910933721A CN110795567A CN 110795567 A CN110795567 A CN 110795567A CN 201910933721 A CN201910933721 A CN 201910933721A CN 110795567 A CN110795567 A CN 110795567A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- data
- module
- user
- platform
- 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.)
- Pending
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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention provides a knowledge graph platform which comprises a data perception layer, a data processing layer, a platform service layer, an application layer and a management layer, wherein the data perception layer collects data, the data processing layer processes the collected data, the processing of the collected data comprises data cleaning and data alignment of the collected data, the platform service layer constructs a knowledge graph model based on the processed data, the application layer provides services for a user based on the constructed knowledge graph model, and the management layer is used for managing the user. The invention provides a knowledge graph platform, which realizes the construction of a knowledge graph model from data acquisition and processing, simultaneously realizes the service provision for users and the user management, realizes the knowledge visualization of the users, and improves the data intelligent management level of the users.
Description
Technical Field
The invention relates to the technical field of knowledge graphs, in particular to a knowledge graph platform.
Background
The generation of the knowledge graph marks that the internet is changed from the traditional document internet to the data internet, and daily websites visited by people are basically an internet oriented to documents and web pages, and in the middle of the internet, main information of the internet is expressed through the web pages, and the convenience of the web pages is easy to understand by human beings, so that the information is convenient to look up at all times, but the knowledge graph has a defect that semantic information is insufficient, so that the machine understanding is difficult. Knowledge maps are born under such solicitations. However, there is no mature knowledge-graph platform in the prior art to provide services to users.
Disclosure of Invention
In order to solve the technical problems, the invention provides a knowledge graph platform, which realizes the construction of a knowledge graph model from data acquisition and processing, simultaneously realizes the service provision for users and the user management, realizes the knowledge visualization of the users, and improves the data intelligent management level of the users.
The invention particularly provides a knowledge graph platform which comprises a data sensing layer, a data processing layer, a platform service layer, an application layer and a management layer, wherein the data sensing layer acquires data, the data processing layer processes the acquired data, the processing of the acquired data comprises data cleaning and data alignment of the acquired data, the platform service layer constructs a knowledge graph model based on the processed data, the application layer provides services for a user based on the constructed knowledge graph model, and the management layer is used for managing the user.
Optionally, the platform service layer includes a knowledge modeling module and an algorithm integration module, the knowledge modeling module is configured to convert processed data into a graph database and construct a knowledge graph model, the algorithm integration module packages various algorithms into callable components by calling through a uniform interface, and the application layer provides services for users based on the graph database and the callable components for packaging various algorithms.
Optionally, the platform service layer includes a knowledge modeling module, an expert modeling module, and an algorithm integration module, the knowledge modeling module is configured to convert processed data into a graph database and construct a knowledge graph model, the expert modeling module is configured to define a specific knowledge model by an expert through a specific language, the algorithm integration module encapsulates various algorithms into callable components by calling through a unified interface, and the application layer provides services for a user based on the graph database, the expert-defined knowledge model, and the callable components encapsulating various algorithms.
Optionally, the knowledge modeling module includes an entity modeling unit and a relationship modeling unit, the entity modeling unit finally forms a modeling label model by selecting data, determining an index, and adding an aggregation function, and generates a graph entity from the processed data, and the relationship modeling unit is configured to associate the graph entities to form a graph database, thereby completing knowledge graph modeling.
Optionally, the application layer includes a knowledge retrieval module, a knowledge identification module, an intelligent simulation module, and a business intelligence module, where the knowledge retrieval module provides a knowledge retrieval service to a user, the knowledge identification module is used to provide a knowledge identification service to the user, the intelligent simulation module is used to provide a simulation service to the user, and the business intelligence module is used to provide a business development service to the user.
Optionally, the knowledge retrieval module provides a knowledge retrieval service to the user, specifically: the user enters a thread and traces all entities and relationships associated with the thread.
Optionally, the knowledge identification module is configured to provide a knowledge identification service to the user, and specifically includes: and the user is assisted in quickly and effectively identifying the data relation by depending on a relevancy algorithm so as to be converted into effective knowledge, enters a relevancy analysis link, selects an identification data source, is analyzed by a training machine, and finally forms a relevancy matrix.
Optionally, the intelligent simulation module is configured to provide a simulation service to a user, and specifically includes: and (4) leading the data to a training machine by the user, and outputting a predictive analysis model by the training machine after continuous training.
Optionally, the business intelligence module is configured to provide a service development service to a user, and specifically includes: based on the knowledge graph, the user directly develops the service facing knowledge.
Optionally, the management layer includes an information management module, a permission management module, a terminal management module, a password management module, a log management module, and a notification module, where the information management module is configured to manage user information, the permission management module is configured to determine an access permission of a user, the terminal management module manages a terminal to which the user accesses, the password management module is configured to manage a password of the user, the log management module is configured to record an operation that the user accesses a platform service and does on the platform, and the notification module is configured to send a message or a notification to all users or a designated user.
Additional features and advantages of the invention will be set forth in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention will be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.
Drawings
The invention is further illustrated by means of the attached drawings, but the embodiments in the drawings do not constitute any limitation to the invention, and for a person skilled in the art, other drawings can be obtained on the basis of the following drawings without inventive effort.
FIG. 1 is a schematic view of the structure of the present invention.
Reference numerals:
the system comprises a data perception layer 1, a data processing layer 2, a platform service layer 3, an application layer 4 and a management layer 5.
Detailed Description
The invention is further described with reference to the following examples.
With reference to fig. 1, an embodiment of the present invention provides a knowledge graph platform, which includes a data sensing layer 1, a data processing layer 2, a platform service layer 3, an application layer 4, and a management layer 5, where the data sensing layer 1 collects data, the data processing layer 2 processes the collected data, the processing of the collected data includes data cleaning and data alignment of the collected data, the platform service layer 3 constructs a knowledge graph model based on the processed data, the application layer 4 provides services to a user based on the constructed knowledge graph model, and the management layer 5 is used to manage the user.
The internet is transformed into a data-oriented internet, information and data in the internet can be understood by a machine, the current enterprise data is similar to the traditional document internet, human beings can understand but can not find the data easily, and machines can find but can not understand easily, so that a new data structure is needed, and the machine can not only quickly identify the data, but also can well understand the data;
the knowledge map platform is mainly oriented to enterprise users, enables enterprises to realize knowledge visualization, intelligent simulation and self-service analysis, enables the enterprises to enhance analysis capability, and assists the enterprises to realize knowledge visualization, self-service graph decision analysis and intelligent application construction by using a data interaction technology, a graph calculation technology and a knowledge reasoning technology, so that the enterprises can comprehensively realize intelligent operation.
The knowledge graph platform has a complete, stable and high-expansibility development system architecture, adopts a B/S architecture, and can be deployed in a virtualized cloud resource environment or a pc-server environment. The platform has high reliability and expansibility, and can continuously work. And the requirements of different enterprises are met. Meanwhile, the platform supports two modes of operation of the micro-service and the virtualization cluster, a complete micro-service operation architecture is formed by relying on a Docker container and a Zookeeper, application multi-activity can be realized through the virtualization cluster such as F5 or Nginx, and in order to simplify operation and maintenance, the platform supports a third-party automatic construction tool to construct and release the application.
The knowledge graph platform realizes the construction of a knowledge graph model from data acquisition and processing, simultaneously realizes the service provision for users and the user management, realizes the knowledge visualization of the users, and improves the data intelligent management level of the users.
Preferably, the platform service layer 3 includes a knowledge modeling module and an algorithm integration module, the knowledge modeling module is used for converting processed data into a graph database and constructing a knowledge graph model, the algorithm integration module packages various algorithms into callable components by calling a uniform interface, and the application layer 4 provides services for users based on the graph database and the callable components for packaging various algorithms.
The algorithm integration module encapsulates various algorithms into a callable component in a unified interface calling mode, the current mainstream algorithm in the knowledge graph platform of the embodiment is realized through Python and encapsulated inside, and in addition, more algorithm injection can be supported in a component expansion mode to meet different service requirements. The platform has cross-language calling capability, and can easily support R language calling and Go language calling.
The algorithm kernel not only encapsulates the algorithm inside, but also provides service to the outside in a micro-service mode, provides machine training capability for the whole application layer 4, and any application passing authentication can call the algorithm service provided by the algorithm kernel in a Restful API mode.
Preferably, the platform service layer 3 includes a knowledge modeling module, an expert modeling module and an algorithm integration module, the knowledge modeling module is used for converting processed data into a graph database and constructing a knowledge graph model, the expert modeling module is used for defining a specific knowledge model through a specific language by an expert, the algorithm integration module encapsulates various algorithms into callable components through a uniform interface calling mode, and the application layer 4 provides services for users based on the graph database, the knowledge model defined by the expert and the callable components encapsulating various algorithms.
In the expert modeling process, business experts can form fixed and solidified problem solutions and troubleshooting solutions according to work experience, and the fixed and solidified problem solutions and troubleshooting solutions are used by common other basic personnel, specifically, the experts define specific models through specific languages, and other personnel use the expert models according to guide steps, and are particularly useful in knowledge retrieval, from the perspective of knowledge used by users, the basic building blocks of knowledge application, from relevant ' problems ', such as ' what is a set of selectable port options sorted by scores given a container, omitted ports and date? With visualization knowledge model management, a user can search concepts, bring them into the workspace, define relationships between them, and apply appropriate functionality to visualize the user's topic expertise.
After the expert models, more people can search knowledge in different categories according to the theme given by the expert and use the knowledge, so that the effective utilization rate of the knowledge is improved, and the use threshold of the knowledge is reduced.
The expert modeling module provides a saved algorithm model which is divided into a main model and other models, wherein the algorithm language comprises Cypher, Python and R language. The user can also add the algorithm model by himself. The model can be checked, modified and deleted, and the use of a user is facilitated.
Preferably, the knowledge modeling module comprises an entity modeling unit and a relationship modeling unit, the entity modeling unit finally forms a modeling label model by selecting data, determining indexes and adding aggregation functions, the processed data is generated into graph entities, and the relationship modeling unit is used for associating the graph entities to form a graph database and complete knowledge graph modeling.
The knowledge modeling module generates the processed data into a graph entity, completes the conversion of the source data to the graph data and the knowledge graph modeling, and provides a bottom support for the subsequent business processing.
The solid modeling unit is used as the most basic and most critical part of knowledge modeling. Selecting data, determining indexes, adding aggregation functions, finally forming a modeling marking model, then generating source data into a final graph entity in a CSV or direct connection data source mode through an entity import mode, providing knowledge entity support for later entity retrieval, and carrying out deletion, query and modification operations on corresponding entities.
The relation modeling unit associates the two graph entities by selecting the starting node and the ending node, and if the two entities complete the import function, the relation modeling unit can complete the relation from the CSV or a mode of directly connecting a data source through a relation import mode. And operations such as viewing, modifying, deleting and the like are provided, and a clearing function is provided for the imported content.
In the knowledge graph modeling process, the establishment of a knowledge graph model can be completed by means of visualization service, data mining and learning, graph storage, association calculation, data governance and integration, data view unification and the like. The visual service provides various service interfaces and supports business scenes, the data governance and integration comprise governance and integration of structured data and unstructured data, the data mining and machine learning comprise natural language processing and full-data machine learning, and the association calculation comprises association calculation based on rules and establishment of relationships among entities.
Since the data of an enterprise is typically distributed among multiple types of files or data sources, this makes it difficult and time consuming to enter a cross-business problem resolution phase. According to the embodiment, multi-source data management is performed in the knowledge graph modeling process, and relevant data can be interactively discovered and safely obtained through the multi-source data management so as to be used in the knowledge graph. The challenge of ever-changing, growing data can also be helped to be solved by being configured to handle repeated or periodic retrieval of the same content. The present embodiment can provide structured data, such as: knowledge modeling of data such as Oracle/DB2/Mysql and the like; unstructured data can be provided, such as: knowledge modeling of XML/JSON and other data.
Preferably, the application layer 4 includes a knowledge retrieval module, a knowledge identification module, an intelligent simulation module and a business intelligence module, the knowledge retrieval module provides a knowledge retrieval service for a user, the knowledge identification module is used for providing a knowledge identification service for the user, the intelligent simulation module is used for providing a simulation service for the user, and the business intelligence module is used for providing a business development service for the user.
The application layer 4 also provides machine training service for users, the machine training takes knowledge maps as a source, and utilizes an algorithm integration module to improve the utilization rate of knowledge, specifically, a platform continuously outputs a stable analysis model by machine training for knowledge map structure and algorithm training, and can also utilize a similarity algorithm to quantify the degree of correlation between data.
The machine training mainly trains a proper analysis model according to a training algorithm through data (historical data or real-time data) given by a user, and finally forms a stable analysis model through continuous training and adjustment, and the stable analysis model is issued to a production area for production prediction analysis, so that production business activities are guided.
The method comprises the steps of obtaining a model name list (adding a model as required) from a database through machine training, obtaining a factor list (adding factors as required) through the same method, uploading CSV files of local samples to a server after determining the model name and the factor name, training through an optional training algorithm (decision tree, RNN, CNN, SVM support vector machine and the like), and downloading and storing remote files.
The knowledge graph platform of the embodiment includes various mainstream algorithms based on scikit learn and tensflow, and specifically includes but is not limited to the following algorithms:
gradient Descent algorithm (Gradient): the basic principle is very simple: searching for a minimum value along the gradient descending direction of the target function (or searching for a maximum value along the gradient ascending direction), more generally speaking, the training process is a process of continuously deriving the function, and once a certain error rate is met, the iteration is stopped, so that the interval tangent rate and the interval adjusting value are solved. Other variant algorithms based on gradient descent algorithms also include: BGD (Batch Gradient Descript), SGD (Stochastic Gradient Descript), MBGD (Mini-Batch Gradient Descript) and some optimization algorithms such as: impulse gradient descent (Momentum), etc., the knowledge graph platform of the embodiment also provides a mode of algorithm late injection for expansion.
And (3) convolutional neural network algorithm: a convolutional network is essentially an input-to-output mapping that is able to learn a large number of input-to-output mapping relationships without any precise mathematical expression between the inputs and outputs, and the network has the ability to map between input-output pairs as long as the convolutional network is trained with known patterns. The convolutional neural network has more layers than the neural network, improves the recognition degree, and provides different improved algorithms for different application scenes by mainly relying on the idea of a hierarchical network in addition to the convolutional neural network and the Recurrent Neural Network (RNN).
Linear regression algorithm: linear regression is based on the assumption that characteristics meet a linear relationship, a model is trained according to given training data, the model is used for prediction, the linear regression is based on the premise that correlation factors meet a simple linear equation which can be a multi-term linear equation, in order to achieve equation fitting, a loss function and a penalty value need to be specified, and a general loss function can pass through a gradient descent equation. In practical application, whether a linear relation exists between data is firstly verified, and interval linear review is carried out in an interval weighted linear regression mode to ensure equation fitting, and the practical situation is probably not simple global linear regression.
Preferably, the knowledge retrieval module provides a knowledge retrieval service to the user, specifically: the user enters a thread and traces all entities and relationships associated with the thread.
The knowledge graph is originally proposed only for improving data retrieval, because many existing data in an enterprise are isolated and unrelated, the data is difficult to utilize, particularly for retrieving and visualizing the enterprise data, programs are basically developed according to needs to be realized, in the past, an enterprise data system is constructed in a pure data oriented or pure text oriented mode, because the knowledge graph appears, the knowledge graph can fundamentally change the knowledge of data application, the knowledge graph is oriented to a data network instead of simple data, and is oriented to specific things (entities) in the data network, the things are not isolated per se, and the things are related with each other, so that the data network of the enterprise is formed. That is to say: firstly, various objects (entities and attributes) exist, then, the complete data network can be formed only by the association relationship among the objects, and in the past, the traditional relational data only retains the information of the data and does not retain the relationship among the data, so that the knowledge retrieval and visualization application is always incomplete, and the relationship context is unclear.
By utilizing the knowledge map platform, enterprises can quickly inquire service information, inquire service relation, quickly position fault information and quickly position information around faults.
Preferably, the knowledge identification module is configured to provide a knowledge identification service to the user, and specifically includes: and the user is assisted in quickly and effectively identifying the data relation by depending on a relevancy algorithm so as to be converted into effective knowledge, enters a relevancy analysis link, selects an identification data source, is analyzed by a training machine, and finally forms a relevancy matrix.
There are many potential associations inherent in enterprise data that are logically difficult to identify as knowledge, relying solely on human mining.
The knowledge graph platform of the embodiment comprises a common correlation coefficient algorithm and supports most of correlation degree identification. Correlation coefficient algorithms include, but are not limited to, the following:
pearson correlation coefficient principle: the data used for describing two linear groups change the trend of movement together, the coefficient changes between-1 and 1, the mathematical formula is expressed as: the pearson correlation coefficient is equal to the covariance of the two variables divided by the standard deviation of the two variables.
Spearman rank correlation coefficient: spearman correlation coefficient, also commonly referred to as spearman rank correlation coefficient. "rank" is understood to mean an order or sequence that is solved according to the sequence position of the original data, and the representation does not have the limitation of solving the correlation coefficient of the pearson.
kendall rank correlation coefficient: the Kendel correlation coefficient, also called Kendel rank correlation coefficient, is also a rank correlation coefficient, but the object calculated by the Kendel correlation coefficient is a classification variable, which can be understood as a categorical variable that can be classified into two categories, namely, unordered and ordered.
Preferably, the intelligent simulation module is configured to provide simulation services to the user, and specifically includes: and (4) leading the data to a training machine by the user, and outputting a predictive analysis model by the training machine after continuous training.
By means of various algorithms integrated by the knowledge graph and the algorithm integration module, a user can perform business simulation deduction through a training machine and intelligent simulation, so that the stability of production batches is guaranteed, and the defective rate of products is reduced. In the traditional statistical analysis method, because a data source is not established on a complete factor pedigree diagram, the production stability can be improved to a certain extent only, and the problem cannot be solved radically, and the historical problem is fundamentally solved by establishing a factor stability analysis model on a knowledge graph.
The intelligent simulation comprises forward simulation and reverse simulation, the forward simulation classifies the trained data according to models and batches, the trained data are displayed in a line graph mode, and a deduction result is obtained by dragging points on the line graph to carry out deduction. The reverse simulation classifies the trained data according to models and batches, displays the data in the form of a histogram, the histogram can select multiple graphic modes at the same time, and obtains a modified deduction result by modifying the result data and deducting the result.
Based on an intelligent simulation result, the knowledge graph platform provides intelligent analysis service, an enterprise can really utilize a knowledge structure to activate business experts, the business experts create and study an analysis model by self, the enterprise is helped to realize business prediction intellectualization, past complete experience is abandoned, and the enterprise is really enabled to enhance analysis capability.
With knowledge map genes, an enterprise can really utilize a knowledge structure to activate business experts, and the business experts create and research an analysis model by self help, so that the enterprise is helped to realize business prediction intellectualization, abandon the past completely depending on experience, and really enable the enterprise to enhance analysis capability. The past BI analysis tools basically rely on underlying source data (database table structure) to perform business intelligent analysis, and since business experts do not know the underlying data structure, self-service analysis is difficult to achieve really, or the analysis can be completed intelligently by developers, so that the current market is fast in change, and the mode can not deal with the fast-changing market.
Preferably, the business intelligence module is used for providing service development services for users, and specifically includes: based on the knowledge graph, the user directly develops the service facing knowledge.
Specifically, the business intelligence established on the knowledge graph platform can really enable enterprise users to carry out innovation analysis by themselves without intervention of developers, and the past business intelligence tools are directly oriented to source data (database tables), and business experts do not understand underlying data structures at all, so that analysis is carried out without next hand, and only the developers can be relied on, and business decision is lagged. Only knowledge (business) oriented business intelligence can really activate the innovation ability of business experts.
Preferably, the management layer 5 includes an information management module, a permission management module, a terminal management module, a password management module, a log management module, and a notification module, where the information management module is configured to manage user information, the permission management module is configured to determine an access permission of a user, the terminal management module is configured to manage a terminal to which the user accesses, the password management module is configured to manage a password of the user, the log management module is configured to record an operation that the user accesses a platform service and does on the platform, and the notification module is configured to send a message or a notification to all users or a designated user.
The information management module is used for managing user information, and specifically comprises:
based on the enterprise sector, the sector reflects the actual organization and business management information. The platform user can add, modify and delete the authority by himself, and can also make and modify the authority range of the department. Each department has a superior department, and a department tree is generated by the departments; each department has information such as full name, abbreviation, number, type, level, etc. The department is associated with a role and a user, the role has a subordinate department, and the subordinate department of the role defines the authority range of the role.
The department role divides the authority responsibility of the users under the department. Department role basis attributes: code, department role name, affiliated department, role name. The platform user can add, modify and delete the department role and can make and modify the authority of the department role.
The user logs on the platform by virtue of a user name and a password. The user has basic information such as the department to which the user belongs, the contact phone and the like. The user has one or more roles, the user authority is the sum of all the role authorities of the user by default, and the authority of the user can be added and deleted within the authority range of departments.
The information management of the user mainly comprises adding a user, modifying the user information, deleting the user and inquiring the user, and can carry out unified management on information such as user name, password, department, contact way and the like.
The authority management module is used for determining the access authority of the user, and specifically comprises the following steps:
the platform authority management can be divided into resource authority, data source authority, SQL table authority and SQL column authority according to types. The specific authority is configured in departments, roles and users.
Platform rights management can be divided into permanent rights and temporary rights by time efficiency. The permanent authority is configured in departments, roles and users. The temporary authority is applied in the temporary authority, and a manager approves and executes the acquisition authority in a task.
The terminal management module manages a terminal accessed by a user, and specifically comprises the following steps:
the platform limits the terminals which are allowed to be accessed by the user, automatically registers the terminals which are accessed by the user, and can log in the terminals through an administrator approval party if the user does not allow the access of the login terminal.
The password management module is used for managing the password of the user, and specifically comprises the following steps:
when a platform user is newly added, a password expiration date is set (within a certain time range), the user needs to modify the user password in the system periodically, and when the user modifies the password, the password expiration date is delayed for a fixed time (configurable). If the user fails to modify the password periodically, the password may expire, so that the user may not log into the system normally.
The log management module is used for recording operations of a user accessing platform services and doing on a platform, and specifically comprises the following steps:
the system records the operations that the user accesses the platform service and does on the platform. The system log function is convenient for users to check and manage the system logs. The method comprises the functions of log grading, log marking, log downloading and the like.
The notification module is used for sending messages or notifications to all users or specified users, and specifically comprises the following steps:
the system notification function in the platform system management can send messages or system notifications to all users or specified users in the system, and the messages or the system notifications comprise sending and viewing of attachments.
The notification message can be viewed, the attachment can be downloaded and the like in the system management and the automatic publishing of the platform.
The management layer 5 also provides data source management, specifically for all machine data source management required by the platform function, and the management is divided into the following types: JDBC, PLT _ FTP, AS400_ FTP, SCP _ FTP, SSH, AS400_ PGM, AS400_ CL, ESB, SFTP, TELNET, WAS, SVN, HTTP, and the like. These data sources implement various functions of the platform by connecting various protocols inside the platform.
The knowledge map platform technology architecture adopts a multilayer structure, and has the advantages of data presentation, service logic, data storage separation, clear codes, convenient development, operation and maintenance personnel division of labor, and rapid and accurate work focus positioning when service changes. On the other hand, the mode can well complete the cooperation task among the businesses, attaches importance to and strengthens the uniformity and normalization of data, and has high flexibility for the expansion of functions.
From the above description of embodiments, it is clear for a person skilled in the art that the embodiments described herein can be implemented in hardware, software, firmware, middleware, code or any appropriate combination thereof. For a hardware implementation, a processor may be implemented in one or more of the following units: an Application Specific Integrated Circuit (ASIC), a Digital Signal Processor (DSP), a Digital Signal Processing Device (DSPD), a Programmable Logic Device (PLD), a Field Programmable Gate Array (FPGA), a processor, a controller, a microcontroller, a microprocessor, other electronic units designed to perform the functions described herein, or a combination thereof. For a software implementation, some or all of the procedures of an embodiment may be performed by a computer program instructing associated hardware. In practice, the program may be stored on or transmitted over as one or more instructions or code on a computer-readable medium. Computer-readable media includes both computer storage media and communication media including any medium that facilitates transfer of a computer program from one place to another. A storage media may be any available media that can be accessed by a computer. Computer-readable media can include, but is not limited to, RAM, ROM, EEPROM, CD-ROM or other optical disk storage, magnetic disk storage or other magnetic storage devices, or any other medium that can be used to carry or store desired program code in the form of instructions or data structures and that can be accessed by a computer.
Finally, it should be noted that the above embodiments are only used for illustrating the technical solutions of the present invention, and not for limiting the protection scope of the present invention, although the present invention is described in detail with reference to the preferred embodiments, it should be understood by the ordinary technical destination in the art that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention.
Claims (10)
1. The knowledge graph platform is characterized by comprising a data perception layer, a data processing layer, a platform service layer, an application layer and a management layer, wherein the data perception layer collects data, the data processing layer processes the collected data, the collected data are processed and include data cleaning and data alignment of the collected data, the platform service layer builds a knowledge graph model based on the processed data, the application layer provides services for a user based on the built knowledge graph model, and the management layer is used for managing the user.
2. The knowledge graph platform of claim 1, wherein the platform service layer comprises a knowledge modeling module and an algorithm integration module, the knowledge modeling module is used for converting processed data into a graph database to construct a knowledge graph model, the algorithm integration module packages various algorithms into callable components by calling through a uniform interface, and the application layer provides services for users based on the graph database and the callable components for packaging various algorithms.
3. The knowledgegraph platform of claim 1, wherein the platform services layer comprises a knowledge modeling module, an expert modeling module, and an algorithm integration module, wherein the knowledge modeling module is configured to convert processed data into a graph database to construct a knowledge graph model, the expert modeling module is configured to allow an expert to define a specific knowledge model in a specific language, the algorithm integration module encapsulates various algorithms into callable components by calling through a uniform interface, and the application layer provides services to a user based on the graph database, the expert-defined knowledge model, and the callable components encapsulating the various algorithms.
4. The knowledge graph platform according to claim 2 or 3, wherein the knowledge modeling module comprises an entity modeling unit and a relationship modeling unit, the entity modeling unit finally forms a modeling label model by selecting data, determining indexes and adding aggregation functions, and generates processed data into graph entities, and the relationship modeling unit is used for associating the graph entities to form a graph database to complete knowledge graph modeling.
5. The knowledgegraph platform of claim 4, wherein the application layer comprises a knowledge retrieval module, a knowledge identification module, an intelligent simulation module, and a business intelligence module, wherein the knowledge retrieval module provides knowledge retrieval services to a user, the knowledge identification module is configured to provide knowledge identification services to the user, the intelligent simulation module is configured to provide simulation services to the user, and the business intelligence module is configured to provide business development services to the user.
6. The knowledge-graph platform of claim 5, wherein the knowledge retrieval module provides knowledge retrieval services to the user, specifically: the user enters a thread and traces all entities and relationships associated with the thread.
7. The knowledge graph platform of claim 5, wherein the knowledge identification module is configured to provide knowledge identification services to the user, specifically: and the user is assisted in quickly and effectively identifying the data relation by depending on a relevancy algorithm so as to be converted into effective knowledge, enters a relevancy analysis link, selects an identification data source, is analyzed by a training machine, and finally forms a relevancy matrix.
8. The knowledgegraph platform of claim 5, wherein the intelligent simulation module is configured to provide simulation services to the user, specifically: and (4) leading the data to a training machine by the user, and outputting a predictive analysis model by the training machine after continuous training.
9. The knowledgegraph platform of claim 5, wherein the business intelligence module is configured to provide business development services to the user, specifically: based on the knowledge graph, the user directly develops the service facing knowledge.
10. The knowledgegraph platform of claim 1, wherein the management layer comprises an information management module, a right management module, a terminal management module, a password management module, a log management module and a notification module, the information management module is used for managing user information, the right management module is used for determining access right of a user, the terminal management module is used for managing a terminal accessed by the user, the password management module is used for managing a password of the user, the log management module is used for recording operations of the user accessing platform services and doing the platform, and the notification module is used for sending messages or notifications to all users or specified users.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910933721.0A CN110795567A (en) | 2019-09-29 | 2019-09-29 | Knowledge graph platform |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910933721.0A CN110795567A (en) | 2019-09-29 | 2019-09-29 | Knowledge graph platform |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110795567A true CN110795567A (en) | 2020-02-14 |
Family
ID=69440011
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910933721.0A Pending CN110795567A (en) | 2019-09-29 | 2019-09-29 | Knowledge graph platform |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110795567A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111737494A (en) * | 2020-06-28 | 2020-10-02 | 上海乂学教育科技有限公司 | Knowledge graph generation method of intelligent learning system |
| CN111930856A (en) * | 2020-07-06 | 2020-11-13 | 北京邮电大学 | Method, device and system for constructing domain knowledge graph ontology and data |
| CN112035622A (en) * | 2020-09-04 | 2020-12-04 | 上海明略人工智能(集团)有限公司 | Integrated platform and method for natural language processing |
| CN112241401A (en) * | 2020-10-16 | 2021-01-19 | 中国民用航空华东地区空中交通管理局 | Knowledge graph-based digital quality management system and method |
| CN112287114A (en) * | 2020-09-28 | 2021-01-29 | 珠海大横琴科技发展有限公司 | Knowledge graph service processing method and device |
| CN113610104A (en) * | 2021-06-29 | 2021-11-05 | 优丹睦风(北京)科技有限公司 | Algorithm management method based on spatial knowledge/model base system |
| CN113918728A (en) * | 2021-09-28 | 2022-01-11 | 安徽国科信通科技有限公司 | Industrial Internet post-service knowledge map analysis platform |
| CN116662569A (en) * | 2023-05-31 | 2023-08-29 | 东北石油大学 | Knowledge-graph-based oil-gas geological evaluation method |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106447346A (en) * | 2016-08-29 | 2017-02-22 | 北京中电普华信息技术有限公司 | Method and system for construction of intelligent electric power customer service system |
| CN106909643A (en) * | 2017-02-20 | 2017-06-30 | 同济大学 | The social media big data motif discovery method of knowledge based collection of illustrative plates |
| CN107341215A (en) * | 2017-06-07 | 2017-11-10 | 北京航空航天大学 | A kind of vertical knowledge mapping classification ensemble querying method of multi-source based on Distributed Computing Platform |
| CN107920266A (en) * | 2017-11-20 | 2018-04-17 | 安徽亿联智能有限公司 | A kind of set-top box users knowledge mapping make |
| US20180121424A1 (en) * | 2016-11-03 | 2018-05-03 | Business Objects Software Limited | Knowledge-driven generation of semantic layer |
| CN108021718A (en) * | 2017-12-29 | 2018-05-11 | 中国电子科技集团公司信息科学研究院 | Capability of Internet of things knowledge mapping and its construction method |
| CN108038136A (en) * | 2017-11-23 | 2018-05-15 | 上海斯睿德信息技术有限公司 | The method for building up and graph inquiring method of Company Knowledge collection of illustrative plates based on graph model |
| CN109446341A (en) * | 2018-10-23 | 2019-03-08 | 国家电网公司 | The construction method and device of knowledge mapping |
| CN110008288A (en) * | 2019-02-19 | 2019-07-12 | 武汉烽火技术服务有限公司 | The construction method in the knowledge mapping library for Analysis of Network Malfunction and its application |
| CN110162642A (en) * | 2019-05-21 | 2019-08-23 | 武汉大学 | Patent knowledge map construction method based on object broker database |
-
2019
- 2019-09-29 CN CN201910933721.0A patent/CN110795567A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106447346A (en) * | 2016-08-29 | 2017-02-22 | 北京中电普华信息技术有限公司 | Method and system for construction of intelligent electric power customer service system |
| US20180121424A1 (en) * | 2016-11-03 | 2018-05-03 | Business Objects Software Limited | Knowledge-driven generation of semantic layer |
| CN106909643A (en) * | 2017-02-20 | 2017-06-30 | 同济大学 | The social media big data motif discovery method of knowledge based collection of illustrative plates |
| CN107341215A (en) * | 2017-06-07 | 2017-11-10 | 北京航空航天大学 | A kind of vertical knowledge mapping classification ensemble querying method of multi-source based on Distributed Computing Platform |
| CN107920266A (en) * | 2017-11-20 | 2018-04-17 | 安徽亿联智能有限公司 | A kind of set-top box users knowledge mapping make |
| CN108038136A (en) * | 2017-11-23 | 2018-05-15 | 上海斯睿德信息技术有限公司 | The method for building up and graph inquiring method of Company Knowledge collection of illustrative plates based on graph model |
| CN108021718A (en) * | 2017-12-29 | 2018-05-11 | 中国电子科技集团公司信息科学研究院 | Capability of Internet of things knowledge mapping and its construction method |
| CN109446341A (en) * | 2018-10-23 | 2019-03-08 | 国家电网公司 | The construction method and device of knowledge mapping |
| CN110008288A (en) * | 2019-02-19 | 2019-07-12 | 武汉烽火技术服务有限公司 | The construction method in the knowledge mapping library for Analysis of Network Malfunction and its application |
| CN110162642A (en) * | 2019-05-21 | 2019-08-23 | 武汉大学 | Patent knowledge map construction method based on object broker database |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111737494A (en) * | 2020-06-28 | 2020-10-02 | 上海乂学教育科技有限公司 | Knowledge graph generation method of intelligent learning system |
| WO2021190668A1 (en) * | 2020-06-28 | 2021-09-30 | 上海松鼠课堂人工智能科技有限公司 | Method for generating knowledge graph of intelligent learning system |
| CN111930856A (en) * | 2020-07-06 | 2020-11-13 | 北京邮电大学 | Method, device and system for constructing domain knowledge graph ontology and data |
| CN112035622A (en) * | 2020-09-04 | 2020-12-04 | 上海明略人工智能(集团)有限公司 | Integrated platform and method for natural language processing |
| CN112287114A (en) * | 2020-09-28 | 2021-01-29 | 珠海大横琴科技发展有限公司 | Knowledge graph service processing method and device |
| CN112241401A (en) * | 2020-10-16 | 2021-01-19 | 中国民用航空华东地区空中交通管理局 | Knowledge graph-based digital quality management system and method |
| CN113610104A (en) * | 2021-06-29 | 2021-11-05 | 优丹睦风(北京)科技有限公司 | Algorithm management method based on spatial knowledge/model base system |
| CN113918728A (en) * | 2021-09-28 | 2022-01-11 | 安徽国科信通科技有限公司 | Industrial Internet post-service knowledge map analysis platform |
| CN116662569A (en) * | 2023-05-31 | 2023-08-29 | 东北石油大学 | Knowledge-graph-based oil-gas geological evaluation method |
| CN116662569B (en) * | 2023-05-31 | 2024-04-16 | 东北石油大学 | Knowledge-graph-based oil-gas geological evaluation method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110795567A (en) | Knowledge graph platform | |
| CN108073686B (en) | Closed-loop unified metadata architecture with universal metadata repository | |
| WO2021103492A1 (en) | Risk prediction method and system for business operations | |
| US20180165604A1 (en) | Systems and methods for automating data science machine learning analytical workflows | |
| US11093501B2 (en) | Searching in a database | |
| Hausenblas et al. | Apache drill: interactive ad-hoc analysis at scale | |
| CN104767813B (en) | Public's row big data service platform based on openstack | |
| CN105122243B (en) | Expansible analysis platform for semi-structured data | |
| CN111712809A (en) | Learning ETL rules by example | |
| US20160035114A1 (en) | Consolidation and customization of graph-based models | |
| WO2019143705A1 (en) | Dimension context propagation techniques for optimizing sql query plans | |
| US10936688B2 (en) | Dynamically enhancing query results | |
| WO2020244023A1 (en) | Information aggregation method and apparatus based on knowledge graph, and device | |
| US20130166602A1 (en) | Cloud-enabled business object modeling | |
| Janković et al. | Schema on read modeling approach as a basis of big data analytics integration in EIS | |
| CN106709017A (en) | Big data-based aid decision making method | |
| US11416514B2 (en) | Interactive dynamic geo-spatial application with enriched map tiles | |
| CN113656647A (en) | Intelligent operation and maintenance oriented engineering archive data management platform, method and system | |
| CN108416524A (en) | Estate planning based on a figure general framework refines deciphering method | |
| CN104933621A (en) | Big data analysis system and method for guarantee ring | |
| Ahsaan et al. | Big data analytics: challenges and technologies | |
| US11741101B2 (en) | Estimating execution time for batch queries | |
| Subramanian et al. | Systems dynamics-based modeling of data warehouse quality | |
| Palaiokrassas et al. | An IoT architecture for personalized recommendations over big data oriented applications | |
| US20230229944A1 (en) | Auto-enriching climate-aware supply chain management |
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 |