CN108268505A - Modeling method and device based on semantic knowledge - Google Patents
Modeling method and device based on semantic knowledge Download PDFInfo
- Publication number
- CN108268505A CN108268505A CN201611262432.5A CN201611262432A CN108268505A CN 108268505 A CN108268505 A CN 108268505A CN 201611262432 A CN201611262432 A CN 201611262432A CN 108268505 A CN108268505 A CN 108268505A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- user
- modeling
- classification
- semantic
- 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
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
- G06F16/35—Clustering; Classification
-
- 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)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The present invention provides modeling method and device based on semantic knowledge, wherein, including：Obtain a knowledge base；It receives and the modeling of cache user acts, user modeling intention is analyzed according to user modeling action, compares user modeling action and current modeling process and inquires the knowledge base, so as to export a recommendation list to user.The present invention has the characteristics that interactive mode and Dynamic recommendation.The present invention is based on semantic knowledge, and semantic knowledge can be learnt and be stored, moreover it is possible to by the way that modeling result is carried out itself extension as new knowledge.Using the present invention, when user modeling, recommendation list will dynamically be sent, this causes modeling process to become more easily and quickly, it may have higher accuracy.
Description
Technical field
The present invention relates to technical field of automation in industry more particularly to a kind of modeling methods and dress based on semantic knowledge
It puts.
Background technology
The attribute of various components, various components in industrial automation a, production system, in process of production
Incidence relation between data and various components that various components generate is intricate, it may for example comprise each during automated production
Incidence relation between a equipment or component and each equipment or component, wherein, component can be engine, speed changer,
Vibration sensing device and roll packer etc., the incidence relation of inter-module can be driving relationship, the vibration sensing of engine and speed changer
Physical couplings of device and roll packer etc..
Modeling tool is widely used to describe industrial automation system very much, is additionally operable to system emulation, describes data and pass
Connection relationship etc..However, most modeling mechanism needs user in depth to understand very much the details of a system, user is also required
With the ability by true system " translation " into a model.Even one, to modeling sophisticated expert, also needs
Many energy are spent to establish a model, and it is also possible to have error.This is because industrial system has a large amount of group
Part and complicated control logic, and error can lead to the mistake in emulation or error diagnosis, and can cause cost increase.
Invention content
First aspect present invention provides the modeling method based on semantic knowledge, wherein, include the following steps：S1 is obtained
One knowledge base；S2, receives and the modeling of cache user acts, and user modeling intention is divided according to user modeling action
Analysis compares user modeling action and current modeling process and inquires the knowledge base, so as to export a recommendation list to user.
The present invention has the characteristics that interactive mode and Dynamic recommendation.The present invention is based on semantic knowledge, and semantic knowledge can be learnt
And store, moreover it is possible to by the way that modeling result is carried out itself extension as new knowledge.Using the present invention, recommend when user modeling
List will dynamically be sent, this causes modeling process to become more easily and quickly, it may have higher accuracy.
Further, the step S1 further includes following steps：It receives and analyzes knowledge input by user, and utilize classification
Being generated with the knowledge extracted has the model of classification information and piece segment information, so as to generate the knowledge base.
Further, the step S1 further includes following steps：Classifying step is had knowledge classification input by user
There are the model of classification information, and one classification of template or model assignment to input；Extraction step extracts knowledge input by user
And generate piece segment information；Step, the probability of the statistics model and described segment information with classification information are calculated, and is generated
Knowledge base.
Further, the knowledge base includes model with classification information and described segment information and described has
The statistical probability of the model of classification information and described segment information.
Further, the knowledge input by user is semantic knowledge.
Further, the semantic knowledge includes basic semantic knowledge, template and library with classification information, sample mould
Type.
Further, the step S2 further includes following steps：Caching step, cache user modeling action；It is intended to analysis
Step, the user modeling for analyzing caching are acted to determine the user modeling intention and user modeling target, and evaluate user
Model the classification of target；Comparison step, compares the current modeling of caching and the classification of user modeling target, and search knowledge base from
And determine matching knowledge；Recommendation step arranges and exports a recommendation list to user.
Further, the modeling method based on semantic knowledge further includes following steps：S3, user modeling complete with
Afterwards, the knowledge base is extended using final mask.
Further, the recommendation list includes any one of following or appoints multinomial：
Classification；
Piece segment information；
Template.
Second aspect of the present invention provides the model building device based on semantic knowledge, wherein, including：One knowledge acquisition mould
Block is used to obtain a knowledge base；One recommending module, is used to receive and the modeling of cache user acts, according to user
User modeling intention is analyzed in modeling action, is compared user modeling action and current modeling process and is inquired the knowledge
User modeling interface is given in library so as to export a recommendation list.The present invention has the characteristics that interactive mode and Dynamic recommendation.This hair
Bright is based on semantic knowledge, and semantic knowledge can be learnt and be stored, moreover it is possible to by by modeling result as new knowledge come into
Row itself extension.Using the present invention, when user modeling, recommendation list will dynamically be sent, this causes modeling process to become more
Easily and quickly, it may have higher accuracy.
Further, the knowledge acquisition module is knowledge analysis module, is used to receive and analyze and input by user knows
Know, and generated using classification and the knowledge extracted with the model of classification information and piece segment information, so as to generate the knowledge
Library.
Further, the knowledge analysis module includes：Sort module is used to obtain knowledge classification input by user
Model with classification information, and one classification of template or model assignment to input；Extraction module is used to extracting user defeated
The knowledge that enters simultaneously generates piece segment information；Computing module is used to count the model with classification information and segment letter
The probability of breath, and generate knowledge base.
Further, the knowledge base includes model with classification information and described segment information and described has
The statistical probability of the model of classification information and described segment information.
Further, the knowledge input by user is semantic knowledge.
Further, the semantic knowledge includes basic semantic knowledge, template and library with classification information, sample mould
Type.
Further, the recommending module further includes：Cache module is used for cache user modeling action；It is intended to analysis
Module, the user modeling for being used to analyze caching are acted to determine the user modeling intention and user modeling target, and assess
Go out the classification of user modeling target；Comparison module is used for the current modeling for comparing caching and the classification of user modeling target, and
Search knowledge base is so that it is determined that matching knowledge；Recommending module is used to arrange and exports a recommendation list to user.
Further, an expansion module is further included, is used for after user modeling is completed, is expanded using final mask
Knowledge base described in exhibition.
Further, the recommendation list includes any one of following or appoints multinomial：
Classification；
Piece segment information；
Template.
Description of the drawings
Fig. 1 is the system architecture diagram according to the modeling method based on semantic knowledge of a specific embodiment of the invention；
Fig. 2 is schematically illustrated to be known according to the basic semantic with classification information of a specific embodiment of the invention
Know；
Fig. 3 schematically illustrates the template of the semantic knowledge according to a specific embodiment of the invention；
Fig. 4 is the recommendation schematic diagram according to the modeling method based on semantic knowledge of a specific embodiment of the invention；
Fig. 5 is the recommendation schematic diagram according to the user interface of a specific embodiment of the invention；
Fig. 6 is the final modeling structure schematic diagram according to a specific embodiment of the invention.
Specific embodiment
Below in conjunction with attached drawing, description of specific embodiments of the present invention.
Modeling mechanism provided by the invention based on semantic knowledge (semantic knowledge) include knowledge input and
User recommends, and the knowledge of user terminal input can be classified with extraction process so as to be stored in knowledge base, then used by analysis
Family is intended to and sends recommendation list.
Fig. 1 is according to the system architecture diagram of the modeling method based on semantic knowledge of a specific embodiment of the invention, such as
System shown in FIG. 1 is divided into two major parts, i.e. user terminal 100 and server end 200 by dotted line, and user terminal has knowledge
Input module 110 and user modeling interface 120.Wherein, user is by being 110 Input knowledge of input module.And user modeling
Interface 120 is an interactive interface (interactive interface), and user can be performed by user's Modeling interface 120
Modeling provides the Dynamic recommendation of modelling component, template and user interface (user operation panel)
(dynamic recommendations), user interface can send user modeling and act to server end 200 in modeling process
And receive the recommendation or response of server end 200.
Semantic knowledge is the semantic model described with some certain semantic standards (semantic standard)
(semantic model), such as RDF, OWL and Modelica.Using semantic knowledge, recommendation list can in user modeling quilt
User is dynamically provided to, this can cause modeling process to become more easily quick, and with more pinpoint accuracy.
Modeling method provided by the invention based on semantic knowledge includes the following steps：
Step S1 is first carried out, obtains a knowledge base K.Wherein, knowledge base K is the execution present invention based on semantic knowledge
Modeling method basis, step S1 be perform the method for the present invention initial step.
Optionally, knowledge base K can be generated according to knowledge input by user.Therefore, the step S1 is preferably also comprised
Following steps：It receives and analyzes knowledge input by user, and generated using classification and the knowledge extracted with classification information
Model and piece segment information, so as to generate the knowledge base K.It should be noted that after knowledge base determines for the first time, above-mentioned steps
It is not just the necessary step for performing the present invention.
Specifically, user is the primary indispensability for performing subsequent recommendation step by 110 Input knowledge of knowledge input module
Condition, the semantic knowledge include the basic semantic knowledge (basic semantic knowledge) with classification information, mould
Plate (template) and library (library), sample pattern (sample model).Wherein, classification and semantic model standard are utilized
(semantic model standards) is constructed and arranged basic semantic knowledge, and the semantic model standard includes ISA-
95th, SSN (Semantic Sensor Ontology) etc..Wherein, template and library are arranged using experience or standard, such as
Modelica libraries etc., the template and library are represented with consistent form.Wherein, it is built using project before
It can be used for serving as the sample of other brand automobile assembly lines with the model of collection sample pattern, such as certain brand automobile of assembly line
This model.User can be sent to the knowledge analysis module of server end 200 by the knowledge that knowledge input module 110 inputs
It 210 and stores.
Wherein, illustratively, in the basic semantic knowledge with classification information as shown in Figure 2, the language in " Core "
Adopted knowledge is made of semantic entity standard (semantic ontology standards), and which depict industrial automations
The common sense (universal knowledge) of system (industrial automation system), including QUDT and
SSN.As shown in Fig. 2, " control system (Control System) ", " processing factory (Process Plant) ", " vehicle
(Vehicle) " and " assembly line (Assembly Line ') " is classification, and relevant semantic criteria and template are assigned to these points
Class.All knowledge input by user, such as the one or more that template and sample pattern will all be assigned to above-mentioned classification.Have
The basic semantic knowledge of classification information is used to identify user view and modeling target.Wherein, ISA-95 is the standard of control system,
It includes vocabulary, definition and relationship positioning.Similarly, ISO-15926 is the standard of processing factory, also includes vocabulary, definition and pass
System's positioning.In addition, engine template is the template of vehicle, driving template is the template of assembly line, is all the mark that engineer establishes
Quasi-mode plate.
Wherein, template shown in Fig. 3 is illustratively " assembly line driving template (Assembly line driving
Template) ", described with semantic criteria RDF.As shown in figure 3, " engine (Motor) ", " gear-box (Gearbox) ",
" friction pulley (Roller) ", " shock sensor (Vibration Sensor) " and " displacement sensor (Displacement
Sensor) " it is building blocks, and the connection between the line expression between above-mentioned building blocks.For example, " engine " and " vibrating sensing
Connection relation between device " and " friction pulley " and " displacement sensor " is connection, " gear-box " respectively and " engine " and
Connection relation between " friction pulley " is driving.
As shown in Figure 1, server end 200 includes knowledge analysis module 210 and recommending module 220.Wherein, the step S1
It further includes classifying step, extraction step and calculates step, therefore analysis module 210 further comprises sort module 212, carries
Modulus block 214 and computing module 216.
Specifically, in the classifying step, knowledge classification input by user is obtained having classification letter by sort module 212
The model of breath, and one classification of template or model assignment to input；In the extraction step, extraction module 212 is used to extract
Knowledge input by user simultaneously generates piece segment information (segment information), is then stored in the segment and described knows
Know in the K of library；In the calculating step, computing module 216 is used to count the model with classification information and the segment
The probability of information, and generate knowledge base.
Then step S2 is performed, receives and the modeling of cache user acts (user actions), moved according to user modeling
Make to analyze user modeling intention, compare user modeling action and currently model and inquire the knowledge base, so as to export
One recommendation list is to user.
As shown in Figure 1, the step S2 further includes caching step, is intended to analytical procedure, comparison step and recommendation step, because
The recommending module 220 of this server end 200 also further cache module 222, be intended to analysis module 224,226 and of comparison module
Recommend execution module 228.
Specifically, in the caching step, cache module 222 acts for cache user modeling, and sends user and build
Mould, which acts, gives intention analysis module 224；In the intention analytical procedure, it is intended that the use that analysis module 224 caches for analysis
Family modeling action evaluates classification and the transmission of user modeling target to determine the user modeling intention and user modeling target
To comparison module；In the comparison step, comparison module 226 is used for the current modeling for comparing caching and user modeling target
Classification, search knowledge base is so that it is determined that matching knowledge and being sent to recommendation execution module 228；In the recommendation step, recommend
Execution module 228 is used to arrange and export a recommendation list to user modeling interface 120, i.e. user.
Wherein, as shown in Figure 1, knowledge base K is the knowledge data base preserved from knowledge analysis module 210, knowledge base K is used for
Recommending module 220 to obtain most matched knowledge compared to user modeling action.The knowledge base K is included with classification information
The statistical probability of model and described segment information and the model and described segment information with classification information.
Model with classification information is from all knowledge organizations input by user, including basic semantic knowledge, mould
Plate and library, sample pattern.As shown in Fig. 2, the model with classification information can be analyzed and preserve to knowledge base K, in addition, user
The knowledge (such as template, library and sample pattern) newly inputted can classify annotation (annotated with classifications)
And it is stored in knowledge.Wherein, template come from two resources, one is template and the knowledge in library input, it is another be sample pattern and
The analysis result of piece segment information.If for example, we cannot to obtain template from " assembly line driving template " as shown in Figure 3 defeated
Enter, then we can analyze sample pattern and piece segment information to obtain building blocks (blocks) and the connection with high likelihood
(connections, more than some building blocks quantity) as a result, for example join probability (Combining Probability) is more than
0.9, a template as shown in Figure 3 can be generated as template knowledge.And difference lies in analyses for template input and analysis template
Template has the ability of description template probability.The above process is performed every time when inputting final mask as sample pattern, because
This, template has the ability of self extension (self expanding).
Segment information includes the connection probability between building blocks, connection and building blocks, the building blocks being also included in different classifications
With the probability of connection.When extracting piece segment information, all users input can all be considered, in fact from some engineering statistic algorithms
(mathematical statistic algorithms) is calculated and is showed with same form.S is represented in Fig. 3
The least model of building blocks including two interconnections, is extracted from input template as shown in Figure 3.Wherein,
As shown in figure 3, piece segment information S includes two building blocks, respectively " engine " and " vibrating sensor ", " engine " and " vibration
Connection relation between sensor " is connection.
1 segment Examples of information of table
| From | To | connection | classification | count | FromProb | ToProb | ConnProb | ClassProb |
| Engine | Gear-box | Driving | Assembly line | 5 | 0.85 | 0.75 | 0.41 | 1.0 |
| Engine | Vibrating sensor | Connection | Assembly line | 6 | 0.76 | 0.66 | 0.62 | 0.96 |
| Gear-box | Friction pulley | Driving | Assembly line | 4 | 0.83 | 0.63 | 0.58 | 0.94 |
| Friction pulley | Displacement sensor | Connection | Assembly line | 2 | 0.64 | 0.88 | 0.49 | 1.0 |
Example of the upper table for a piece segment information, as shown above, " from " and " to " represents building blocks and connection in segment
Direction, schematically illustrate such as the connection relation between the building blocks representated by the arrow direction of Fig. 3.
" classification " is the sort module 212 in knowledge analysis module 210 as shown in Figure 2 to analyze and preserve.
" count " represents the frequency that the segment occurs." FromProb ", " ToProb ", " ConnProb " and " ClassProb " is each
The calculating of element conditional probability (conditional probabilities).
It is further described by taking segment S in Fig. 3 as an example below.Wherein, conditional probability P (B | A) represents to have occurred and that as A
When B probability, wherein, A and B represent different building blocks.Therefore, we can obtain：
FromProb=P (motor-driven gear case | engine)=0.85；
ToProb=P (motor-driven gear case | gear-box)=0.75；
ConnProb=P (motor-driven gear case | driving)=0.41；
ClassProb=P (assembly line | motor-driven gear case)=1.0
Due to different fields (domain) and usage (usage), all piece segment informations can be classified annotation.If I
To be obtained from processing model analysis (analysis of processing model) as a result, this will be particularly useful.Due to
There is the model of maximum probability in classification range, therefore recommendation selection meeting in these classification and its accurately.
Further, actions of the step S2 based on user and knowledge base are recommended, and step S2 is action drives process
(action triggered process).The purpose of step S2 is in order to analyze user modeling action and knowledge base, to find
Most matched knowledge, and it is supplied to user's the most useful suggestion in modeling process.Fig. 5 is according to one specific implementation of the present invention
The recommendation schematic diagram of the modeling method based on semantic knowledge of example, as shown in figure 5, arrow 1~5 indicates different operations.Arrow
First 1 represents that modeling action or recommendation selection can be sent to when user performs modeling action or has selected a kind of recommendation
Cache module 222.Arrow 2 represents modeling action or selection is recommended to be buffered.Arrow 3 represents that being intended to analysis module 224 can analyze
The modeling action of caching recommends selection to determine the classification of user view.When user selects one in a recommendation list,
When the selection belong to classification A, it means that user more likely in the A that classify model rather than other options belonging to classification.
Arrow 4 represents that comparison module 226 can combine action and generate modeling, can utilize the modeling and classification being intended in analysis module 224
Compare with knowledge base K, and obtain most matched knowledge to export comparison result, comparison result include classification, piece segment information and
Template.Arrow 5 represents to recommend the meeting tissue comparison result of execution module 228 to a recommendation list and sends recommendation list to user
Modeling interface 120, recommendation list include classification, piece segment information and template.
Wherein, the classification in recommendation results is generated from intention analysis module 224 or comparison module 226.Work as modeling
In building blocks have belong to a certain classification Maximum Possibility when, which can then be recommended.Segment is in knowledge based library
Piece segment information is come what is calculated, and when user's execution acts, the segment for belonging to operation building blocks can be collected and calculated with probability, from
And list ranking.Modeling module can be used in same classification compared with template, and calculate matching rate, so as to the row of listing
Name.
Fig. 5 is according to the recommendation schematic diagram of the user interface of a specific embodiment of the invention, in interface 1201, user
It has been selected in interface one " engine ", recommendation gives a tabulation, which includes " assembly line ", " vehicle
" and " processing factory " for user select.Then user modeling interface 120 is switched to interface 1202 from interface 1201, is held in user
After row modeling action, user has selected " gear-box " and " friction pulley " to establish such as interface 1203 from the correlation type of recommendation
Shown model.At this point, a recommendation list is recommended again in interface 1203, including relevant classification and template.The phase
It closes classification and includes " shock sensor " and " displacement sensor ", the template includes " assembly line driving template ".
Finally, step S3 is performed, after user modeling is completed, is known using described in final mask (final model) extension
Know library.Final mask is sent to knowledge analysis module 210 to extend knowledge base K, and classification, segment and template can quilts in this step
It rearranges, the probability of classification, segment and template can be recalculated.As shown in fig. 6, segment can be extracted from model, newly
Segment such as " slide plate line " and " track " and their probability can be calculated and add in knowledge base K.It is as shown in fig. 6, final
Modeling tool there are two template as shown in Figure 3, wherein, 1 sliding tooth roller box 1 of engine, gear-box 1 drives friction pulley 1, friction
1 connection displacement sensor 1 of wheel, engine 1 connect vibrating sensor.Similarly, 2 sliding tooth roller box 2 of engine, gear-box 2 drive
Friction pulley 2, friction pulley 2 connect displacement sensor 2, and engine 2 connects vibrating sensor.Wherein, engine 1 is close to engine 2.
For the relationship of slide plate line and track to there is track, track and the relationship of engine 1 and engine 2 are respectively to have driving.
In addition, existing segment and template can be analyzed based on new knowledge base.Piece with similary diagram (schema)
Section and template can be reinforced, and opposite part can be retrieved, therefore knowledge base K has obtained itself extension (self-
Expanding) and itself strengthens (self-enhancing).
Most of modeling mechanism of the prior art needs user to understand each details of system in depth, and needing to have will be true
System translates into the ability and technical ability of model.
Second aspect of the present invention additionally provides a kind of model building device based on semantic knowledge, wherein, including：
One knowledge acquisition module is used to obtain a knowledge base K；
One recommending module 220, is used to receive and the modeling of cache user acts, and is acted according to user modeling to user
Modeling is intended to be analyzed, and compares user modeling action and current modeling process and inquires the knowledge base K, so as to export one
Recommendation list is to user modeling interface 120.
Further, the knowledge acquisition module is knowledge analysis module 210, is used to receive and analyze input by user
Knowledge, and being generated using classification and knowledge extract with the model of classification information and piece segment information, so as to know described in generating
Know library K.
Further, the knowledge analysis module 210 includes：
Sort module 212 is used to knowledge classification input by user is obtained the model with classification information, and to input
Template or model assign one classification；
Extraction module 214 is used to extract knowledge input by user and generates piece segment information；
Computing module 216 is used to count the probability of the model with classification information and described segment information, and raw
Into knowledge base K.
Further, the knowledge base K includes model with classification information and described segment information and described has
The statistical probability of the model of classification information and described segment information.
Further, the knowledge input by user is semantic knowledge.
Further, the semantic knowledge includes basic semantic knowledge, template and library with classification information, sample mould
Type.
Further, the recommending module 220 further includes：
Cache module 222 is used for cache user modeling action；
Be intended to analysis module 224, be used to analyzing caching user modeling act with determine the user modeling be intended to and
User modeling target, and evaluate the classification of user modeling target；
Comparison module 226 is used for the current modeling for comparing caching and the classification of user modeling target, and search knowledge base
K is so that it is determined that matching knowledge；
Recommending module 228 is used to arrange and exports a recommendation list to user.
Further, the invention also includes an expansion modules, are used for after user modeling is completed, utilize final mould
Type extends the knowledge base.
Further, the recommendation list includes any one of following or appoints multinomial：
Classification；
Piece segment information；
Template.
Since the model building device provided by the invention based on semantic knowledge and method correspond, above to being based on language
The modeling method of adopted knowledge is described in detail, and herein for simplicity, repeats no more.
The present invention has the characteristics that interactive mode and Dynamic recommendation.The present invention is based on semantic knowledge, semantic knowledge energy
It is enough to be learnt and stored, moreover it is possible to by the way that modeling result is carried out itself extension as new knowledge.Using the present invention, when user builds
Recommendation list will dynamically be sent during mould, this causes modeling process to become more easily and quickly, it may have higher accuracy.
Although present disclosure is discussed in detail by above preferred embodiment, but it should be appreciated that above-mentioned
Description is not considered as limitation of the present invention.After those skilled in the art have read the above, for the present invention's
A variety of modifications and substitutions all will be apparent.Therefore, protection scope of the present invention should be limited to the appended claims.
In addition, any reference numeral in claim should not be considered as to the involved claim of limitation；One word of " comprising " is not excluded for
Unlisted device or step in other claims or specification；The words such as " first ", " second " are only used to indicate names, and
It does not represent any particular order.
Claims (18)
1. the modeling method based on semantic knowledge, wherein, include the following steps：
S1 obtains a knowledge base；
S2, receives and the modeling of cache user acts, and user modeling intention is analyzed according to user modeling action, compares use
Family modeling action and current modeling process simultaneously inquire the knowledge base, so as to export a recommendation list to user.
2. the modeling method according to claim 1 based on semantic knowledge, which is characterized in that the step S1 further include as
Lower step：
Receive and analyze knowledge input by user, and generated using classification and knowledge extract have the model of classification information with
Piece segment information, so as to generate the knowledge base.
3. the modeling method according to claim 2 based on semantic knowledge, which is characterized in that the step S1 further include as
Lower step：
Knowledge classification input by user is obtained the model with classification information, and to the template or model of input by classifying step
Assign a classification；
Extraction step extracts knowledge input by user and generates piece segment information；
Step, the probability of the statistics model and described segment information with classification information are calculated, and generates knowledge base.
4. the modeling method according to claim 3 based on semantic knowledge, which is characterized in that the knowledge base includes having
The statistics of the model of classification information and described segment information and the model and described segment information with classification information is general
Rate.
5. the modeling method according to claim 2 based on semantic knowledge, which is characterized in that the knowledge input by user
For semantic knowledge.
6. the modeling method according to claim 5 based on semantic knowledge, which is characterized in that the semantic knowledge includes tool
There are basic semantic knowledge, template and library, the sample pattern of classification information.
7. the modeling method according to claim 1 based on semantic knowledge, which is characterized in that the step S2 further include as
Lower step：
Caching step, cache user modeling action；
It is intended to analytical procedure, the user modeling for analyzing caching is acted to determine the user modeling intention and user modeling target,
And evaluate the classification of user modeling target；
Comparison step compares the current modeling of caching and the classification of user modeling target, and search knowledge base is so that it is determined that matching
Knowledge；
Recommendation step arranges and exports a recommendation list to user.
8. the modeling method according to claim 1 based on semantic knowledge, which is characterized in that described based on semantic knowledge
Modeling method further includes following steps：
S3 after user modeling is completed, extends the knowledge base using final mask.
9. the modeling method according to claim 1 based on semantic knowledge, which is characterized in that the recommendation list include with
Lower any one is appointed multinomial：
Classification；
Piece segment information；
Template.
10. the model building device based on semantic knowledge, wherein, including：
One knowledge acquisition module is used to obtain a knowledge base (K)；
One recommending module (220), is used to receive and the modeling of cache user acts, and user is built according to user modeling action
Mould intention is analyzed, and is compared user modeling action and current modeling process and is inquired the knowledge base (K), so as to export one
Recommendation list gives user modeling interface (120).
11. the model building device according to claim 10 based on semantic knowledge, which is characterized in that the knowledge acquisition module
It for knowledge analysis module (210), is used to receive and analyzes knowledge input by user, and produced using classification and the knowledge extracted
It is raw that there is the model of classification information and piece segment information, so as to generate the knowledge base (K).
12. the model building device according to claim 11 based on semantic knowledge, which is characterized in that the knowledge analysis module
(210) include：
Sort module (212) is used to knowledge classification input by user is obtained the model with classification information, and to input
Template or model assign a classification；
Extraction module (214) is used to extract knowledge input by user and generates piece segment information；
Computing module (216), is used to count the probability of the model with classification information and described segment information, and generates
Knowledge base (K).
13. the model building device according to claim 12 based on semantic knowledge, which is characterized in that knowledge base (K) packet
Include model with classification information and described segment information and the model with classification information and described segment information
Statistical probability.
14. the model building device according to claim 11 based on semantic knowledge, which is characterized in that described input by user to know
Know for semantic knowledge.
15. the model building device according to claim 14 based on semantic knowledge, which is characterized in that the semantic knowledge includes
Basic semantic knowledge, template and library with classification information, sample pattern.
16. the model building device according to claim 10 based on semantic knowledge, which is characterized in that the recommending module
(220) it further includes：
Cache module (222) is used for cache user modeling action；
It is intended to analysis module (224), the user modeling action for being used to analyze caching is intended to and is used with the determining user modeling
Family models target, and evaluates the classification of user modeling target；
Comparison module (226) is used for the current modeling for comparing caching and the classification of user modeling target, and search knowledge base
(K) so that it is determined that matching knowledge；
Recommending module (228) is used to arrange and exports a recommendation list to user.
17. the model building device according to claim 10 based on semantic knowledge, which is characterized in that it further includes an extension
Module is used for after user modeling is completed, and extends the knowledge base using final mask.
18. the model building device according to claim 10 based on semantic knowledge, which is characterized in that the recommendation list includes
Following any one is appointed multinomial：
Classification；
Piece segment information；
Template.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611262432.5A CN108268505A (en) | 2016-12-30 | 2016-12-30 | Modeling method and device based on semantic knowledge |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611262432.5A CN108268505A (en) | 2016-12-30 | 2016-12-30 | Modeling method and device based on semantic knowledge |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108268505A true CN108268505A (en) | 2018-07-10 |
Family
ID=62755213
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611262432.5A Pending CN108268505A (en) | 2016-12-30 | 2016-12-30 | Modeling method and device based on semantic knowledge |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108268505A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020124377A1 (en) * | 2018-12-18 | 2020-06-25 | 西门子股份公司 | Modeling knowledge acquisition method and apparatus, and system |
| CN112148959A (en) * | 2019-06-27 | 2020-12-29 | 百度在线网络技术（北京）有限公司 | Information recommendation method and device |
| CN114579821A (en) * | 2021-12-13 | 2022-06-03 | 北京市建筑设计研究院有限公司 | Method and device for pushing modeling tool, electronic device and storage medium |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102096717A (en) * | 2011-02-15 | 2011-06-15 | 百度在线网络技术（北京）有限公司 | Search method and search engine |
| CN103440309A (en) * | 2013-08-26 | 2013-12-11 | 胡宝清 | Automatic resource and environment model combination modeling semantic recognition and recommendation method |
| US8700624B1 (en) * | 2010-08-18 | 2014-04-15 | Semantifi, Inc. | Collaborative search apps platform for web search |
| CN104809186A (en) * | 2015-04-20 | 2015-07-29 | 广东工业大学 | Constructing method for mold design and manufacturing knowledge base |
| CN105989097A (en) * | 2015-02-12 | 2016-10-05 | 北京仿真中心 | Ontology-based knowledge base query method and system |
-
2016
- 2016-12-30 CN CN201611262432.5A patent/CN108268505A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8700624B1 (en) * | 2010-08-18 | 2014-04-15 | Semantifi, Inc. | Collaborative search apps platform for web search |
| CN102096717A (en) * | 2011-02-15 | 2011-06-15 | 百度在线网络技术（北京）有限公司 | Search method and search engine |
| CN103440309A (en) * | 2013-08-26 | 2013-12-11 | 胡宝清 | Automatic resource and environment model combination modeling semantic recognition and recommendation method |
| CN105989097A (en) * | 2015-02-12 | 2016-10-05 | 北京仿真中心 | Ontology-based knowledge base query method and system |
| CN104809186A (en) * | 2015-04-20 | 2015-07-29 | 广东工业大学 | Constructing method for mold design and manufacturing knowledge base |
Non-Patent Citations (1)
| Title |
|---|
| 麦强 等: ""基于本体语义建模的航天工程风险管理知识库"", 《图书馆学刊》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020124377A1 (en) * | 2018-12-18 | 2020-06-25 | 西门子股份公司 | Modeling knowledge acquisition method and apparatus, and system |
| CN113168283A (en) * | 2018-12-18 | 2021-07-23 | 西门子股份公司 | Knowledge acquisition method, device and system for modeling |
| CN112148959A (en) * | 2019-06-27 | 2020-12-29 | 百度在线网络技术（北京）有限公司 | Information recommendation method and device |
| CN114579821A (en) * | 2021-12-13 | 2022-06-03 | 北京市建筑设计研究院有限公司 | Method and device for pushing modeling tool, electronic device and storage medium |
| CN114579821B (en) * | 2021-12-13 | 2023-04-11 | 北京市建筑设计研究院有限公司 | Method and device for pushing modeling tool, electronic device and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| JP4239932B2 (en) | production management system | |
| Lease | On quality control and machine learning in crowdsourcing | |
| US20180300333A1 (en) | Feature subset selection and ranking | |
| CN105701120A (en) | Method and apparatus for determining semantic matching degree | |
| CN108268505A (en) | Modeling method and device based on semantic knowledge | |
| Navinchandran et al. | Discovering critical KPI factors from natural language in maintenance work orders | |
| US10460240B2 (en) | Apparatus and method for tag mapping with industrial machines | |
| CN116882978B (en) | Deep sea submersible operation and maintenance support system based on product information frame | |
| WO2020162884A1 (en) | Parameter suggestion system | |
| Gönnheimer et al. | Automated identification of parameters in control systems of machine tools | |
| KR102411291B1 (en) | Method of evaluating quality of smart factory data | |
| CN107977454A (en) | The method, apparatus and computer-readable recording medium of bilingual corpora cleaning | |
| Gunawan et al. | C4. 5, K-Nearest Neighbor, Naïve Bayes, and Random Forest Algorithms Comparison to Predict Students' on TIME Graduation | |
| CN117519656A (en) | Software development system based on intelligent manufacturing | |
| Ziegenbein et al. | Machine learning algorithms in machining: A guideline for efficient algorithm selection | |
| CN116910650A (en) | Data identification method, device, storage medium and computer equipment | |
| Yang | Simultaneous monitoring of mean vector and covariance matrix shifts in bivariate manufacturing processes using hybrid ensemble learning-based model | |
| Lindemann et al. | Methodical data-driven integration of perceived quality into the product development process | |
| CN110659266A (en) | Data processing method of model | |
| Trinks | A classification of real time analytics methods. an outlook for the use within the smart factory | |
| Abdulrahman et al. | An Overview of the Algorithm Selection Problem | |
| Mayilvaganan et al. | Recent trends in human computer interface to analysis the cognitive skill of students based on user interface | |
| Amayri et al. | A statistical process control chart approach for occupancy estimation in smart buildings | |
| Eleftheriadis et al. | The importance of key performance indicators that can contribute to autonomous quality control | |
| CN116975447B (en) | Portable learning machine and resource retrieval method thereof |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180710 |
|
| WD01 | Invention patent application deemed withdrawn after publication |