CN117908637A - Legal knowledge retrieval device based on knowledge graph - Google Patents
Legal knowledge retrieval device based on knowledge graph Download PDFInfo
- Publication number
- CN117908637A CN117908637A CN202410067417.3A CN202410067417A CN117908637A CN 117908637 A CN117908637 A CN 117908637A CN 202410067417 A CN202410067417 A CN 202410067417A CN 117908637 A CN117908637 A CN 117908637A
- Authority
- CN
- China
- Prior art keywords
- knowledge
- data
- module
- graph
- legal
- 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
- 238000010276 construction Methods 0.000 claims abstract description 53
- 230000000007 visual effect Effects 0.000 claims abstract description 17
- 238000000034 method Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 claims abstract description 7
- 230000001502 supplementing effect Effects 0.000 claims abstract description 4
- 238000012545 processing Methods 0.000 claims description 38
- 238000004458 analytical method Methods 0.000 claims description 30
- 230000004927 fusion Effects 0.000 claims description 24
- 238000007405 data analysis Methods 0.000 claims description 19
- 238000003860 storage Methods 0.000 claims description 14
- 230000014509 gene expression Effects 0.000 claims description 6
- 230000010354 integration Effects 0.000 claims description 6
- 238000005457 optimization Methods 0.000 claims description 4
- 238000007781 pre-processing Methods 0.000 claims description 4
- 238000013441 quality evaluation Methods 0.000 claims description 4
- 238000012800 visualization Methods 0.000 claims description 4
- 238000009826 distribution Methods 0.000 claims description 3
- 230000003993 interaction Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 3
- 238000013480 data collection Methods 0.000 description 6
- 238000007726 management method Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 4
- 238000013461 design Methods 0.000 description 4
- 230000006870 function Effects 0.000 description 4
- 238000012360 testing method Methods 0.000 description 4
- 239000013598 vector Substances 0.000 description 4
- 238000013499 data model Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000004140 cleaning Methods 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000003058 natural language processing Methods 0.000 description 2
- 241000255925 Diptera Species 0.000 description 1
- 206010043417 Therapeutic response unexpected Diseases 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 230000002068 genetic effect Effects 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000035772 mutation Effects 0.000 description 1
- 238000007500 overflow downdraw method Methods 0.000 description 1
- 108090000623 proteins and genes Proteins 0.000 description 1
- 230000006798 recombination Effects 0.000 description 1
- 238000005215 recombination Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 238000012546 transfer Methods 0.000 description 1
Landscapes
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses a legal knowledge retrieval device based on a knowledge graph, which comprises the following components: the device comprises a device main body, a touch screen and wheels; the searching module is used for searching related legal knowledge by a user through touch control; the map construction module is used for constructing the knowledge map before providing search and continuously updating the map in the process of search; the internal database is used for storing related information of the established knowledge graph; the knowledge reasoning module is used for supplementing and perfecting the newly established knowledge graph; and the visual display module is used for visually displaying the searched content and simultaneously visualizing the map for further searching. According to the invention, knowledge is retrieved by constructing the knowledge graph, and the knowledge graph is updated continuously, so that the knowledge is acquired more accurately and more effectively through the knowledge graph, and the visual display module is set, and the display effect is optimized.
Description
Technical Field
The invention relates to the technical field of knowledge retrieval, in particular to a legal knowledge retrieval device based on a knowledge graph.
Background
Knowledge graph, one of the most important techniques of current natural language processing technology, was originally proposed by Gogle to enhance search engines, which were then considered to be a massive semantic network containing entities, concepts, and semantic relationships. The application range of the knowledge graph can be divided into a general knowledge graph and a domain knowledge graph, wherein the general knowledge graph has the characteristics of coarse granularity, wide coverage, large volume and the like, and is more well known as YAGO, DBPedia, microsoft Concept Graph and the like. YAGO combines WordNet and Wikipedia knowledge to form a large-scale knowledge base containing characters, cities, countries, movies and organizations, and aims at having more than 1000 ten thousand people of entities, and the relation triplets discuss 45 hundred million people. DBpedia is constructed by extracting structured information from wikipedia information by using specific rules, and approximately contains 500 ten thousand entities, 580 ten thousand triples, microsoft Concept Graph is a large-scale knowledge graph created by Microsoft, and data of DBpedia mainly originate from hundreds of millions of web pages and accumulated search logs, so that the accuracy of a natural language processing scene can be improved, and the DBpedia is applied to the fields of advertising, searching, recommending and the like. The domain knowledge graph generally has the advantages of fine granularity, high accuracy and strong specialization.
At present, when legal knowledge learning is performed, related information can be acquired in few ways, meanwhile, the acquired related information may not be accurate enough, the problem can be effectively solved by utilizing a knowledge graph to perform legal knowledge acquisition, and meanwhile, the knowledge graph is continuously updated along with the increase of searching times, so that the accuracy is effectively improved.
Disclosure of Invention
The invention aims to solve the defect that related information can be acquired in a few ways when legal knowledge learning is performed in the prior art, and the acquired related information may not be accurate enough.
In order to achieve the above purpose, the present invention adopts the following technical scheme:
a legal knowledge retrieval device based on a knowledge graph, comprising:
a device main body for supporting other components as an entire main body of the device;
The touch screen is used for interaction between a user and a system connected with the device;
wheels for moving the device and for fixing the position;
The searching module is used for searching related legal knowledge by a user through touch control;
The map construction module is used for constructing the knowledge map before providing search and continuously updating the map in the process of search;
The internal database is used for storing related information of the established knowledge graph;
the knowledge reasoning module is used for supplementing and perfecting the newly established knowledge graph;
the visual display module is used for visually displaying the searched content and simultaneously visualizing the map for further searching;
The device comprises a device body, wherein a touch screen is arranged outside the device body, wheels are arranged outside the device body, the touch screen obtains touch information after a user touches the device body and transmits the information to a search module, the search module sends a data calling request to an internal database, the internal database responds to a signal after receiving the request signal, the internal database calls legal knowledge related to search content and transmits the legal knowledge to the touch screen for display, the internal database obtains and stores a constructed knowledge graph from a graph construction module, the graph construction module transmits the constructed preliminary graph to a knowledge reasoning module, the knowledge reasoning module transmits the supplemented knowledge graph to a graph construction module, the graph construction module transmits data to a visual display module, and the visual display module transmits the data to the internal database.
The technical scheme further comprises the following steps:
The touch screen transmits touch information to the information interface, the information interface transmits data to the input analysis module, the input analysis module transmits the analyzed data to the search module, the information interface application receives information transmitted by the touch screen and transmits the information to the system, the input analysis module is used for analyzing the data transmitted by the touch screen and performing text analysis on text information input by a user, and the search module acquires the text information transmitted by the input analysis and retrieves the information.
The map construction module further comprises a knowledge fusion unit, a knowledge processing unit, a knowledge map construction unit and a knowledge storage unit, wherein the knowledge fusion unit transmits data to the knowledge processing unit, the knowledge processing unit transmits data to the knowledge map construction unit, the knowledge map construction unit transmits data to the knowledge storage unit, the map updating unit transmits data to the knowledge fusion unit, the knowledge fusion unit is used for integrating new knowledge to eliminate contradiction and ambiguity, for example, a certain object has various expressions, a certain description represents a plurality of objects, at the moment, the knowledge fusion unit performs integration processing on the knowledge fusion data, the knowledge processing unit adds qualified parts into a database through quality evaluation, the knowledge map construction unit is used for constructing a relation between knowledge, and the knowledge storage unit is used for storing the knowledge blocks and transmitting the data.
The system comprises a map construction module, a data analysis module, a data processing module and a data collection module, wherein knowledge data acquired by the map construction module are from a legal knowledge base, the legal knowledge base receives a data request of the data collection module, relevant knowledge is acquired and is transmitted to the data collection module, the data collection module transmits collected data to the data analysis module, the data analysis module transmits data obtained through analysis to the data processing module, the data processing module transmits the data to the map construction module, the legal knowledge base is used for storing relevant knowledge of laws and is communicated with big data, a large number of detailed legal knowledge resources are stored, the data collection module is used for collecting required legal knowledge, the data analysis module is used for carrying out data analysis on the collected data, carrying out category classification and relevant relation analysis on the data, and the data processing module is used for preprocessing the data and providing assistance for integration of subsequent data.
The visual display module displays the knowledge graph, mainly provides the relation between the entities, the related knowledge in the current field and the like for the user, for example, when the national fire is searched, the knowledge related to the national fire is displayed, and meanwhile, the related knowledge is marked with the eye-catching mark, so that the user can acquire legal knowledge conveniently.
The knowledge graph construction unit is used for linking the knowledge with the keywords, and during data analysis, firstly judging the searched content, judging which legal knowledge is needed for the content, if search distribution is carried out, firstly judging which legal knowledge relates to the respective problems, and meanwhile summarizing and classifying the knowledge, so that an intuitive knowledge acquisition way is provided for a user.
The visualization module is provided with a personalized management and typesetting optimization module for displaying the map, the interface is optimized through personalized management, the user can conveniently find and acquire information, meanwhile, in the aspect of typesetting, an administrator optimizes typesetting according to user feedback, if certain inquiry or knowledge acquisition links are not easy to find or display content typesetting is not visual enough, and the administrator can optimize according to reality.
When the user uses the map updating unit each time, the collected information is connected with the search word and transmitted to the knowledge fusion unit to reestablish and update the map, and the map updating unit adds knowledge relations for the map along with the increase of the use times of the user, so that the user can acquire the knowledge more conveniently.
The invention has the following beneficial effects:
1. According to the invention, the knowledge is retrieved by constructing the knowledge graph, and the knowledge graph is updated continuously, so that the knowledge acquisition is more accurate and more effective through the knowledge graph.
2. In the invention, the visual display module is set, and the display effect is optimized.
Drawings
Fig. 1 is a schematic structural diagram of a legal knowledge retrieval device based on a knowledge graph;
FIG. 2 is a system block diagram of a legal knowledge retrieval device based on a knowledge graph;
FIG. 3 is a system block diagram of a map construction module in the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
As shown in fig. 1, the legal knowledge retrieval device based on the knowledge graph provided by the invention comprises:
A device main body 1 for supporting other components as an entire main body of the device;
the touch screen 2 is used for interaction between a user and a system connected with the device;
wheels 3 for moving the device and for fixing the position;
The searching module is used for searching related legal knowledge by a user through touch control;
The map construction module is used for constructing the knowledge map before providing search and continuously updating the map in the process of search;
The internal database is used for storing related information of the established knowledge graph;
the knowledge reasoning module is used for supplementing and perfecting the newly established knowledge graph;
the visual display module is used for visually displaying the searched content and simultaneously visualizing the map for further searching;
The device comprises a device main body 1, a touch screen 2, wheels 3, a searching module, a data calling request, an internal database, a visual display module and a map construction module, wherein the touch screen 2 is arranged outside the device main body 1, the wheels 3 are arranged outside the device main body 1, the touch screen obtains touch information after being touched by a user and transmits the information to the searching module, the searching module sends a data calling request to the internal database, the internal database responds to the signal after receiving the request signal, the internal database calls legal knowledge related to search content and transmits the legal knowledge to the touch screen for display, the internal database obtains and stores a constructed knowledge map from the map construction module, the map construction module transmits the constructed preliminary map to the knowledge reasoning module, the knowledge reasoning module transmits the supplemented knowledge map to the map construction module, the map construction module transmits the data to the visual display module and transmits the data to the internal database.
The touch screen 2 transmits touch information to the information interface, the information interface transmits data to the input analysis module, the input analysis module transmits the data obtained by analysis to the search module, the information interface receives information transmitted by the touch screen and transmits the information to the system, the input analysis module is used for analyzing the data transmitted by the touch screen, text analysis is carried out on text information input by a user, and the search module acquires the text information transmitted by the input analysis and retrieves the information.
In the embodiment of the invention, the device main body 1 is arranged in each cell, so that people can conveniently know and use the device, the wheels 3 can be locked when the device is not moved, a user can use the device through the touch screen 2, search for related knowledge is carried out by touching the touch screen, text input is carried out on the touch screen, the content to be searched is input, the device searches based on a knowledge graph according to the input content and displays the related knowledge, the displayed content not only comprises the related knowledge, but also displays the connection between the knowledge, and when the user needs to move, the user can move through the wheels 3.
According to the characteristics of the system and the device, the connection is carried out by adopting wired connection (such as USB, HDMI, network cable and the like) and wireless connection (such as Wi-Fi, bluetooth and the like).
Wired connection, preparing corresponding cable or adapter; and (3) wireless connection, and preparing a corresponding wireless network card or a Bluetooth adapter.
And determining a connection mode according to the requirements of the system and the device. The system and the device perform one-time pairing or setting, so that smooth connection is ensured.
The connection operation is performed according to the specification or direction of the connection device. The cable is inserted into the corresponding interface through the wired connection; the wireless connection performs some additional settings such as entering a password or performing a pairing operation.
After the connection is completed, a test is performed to ensure that the connection is successful. The test may be performed by testing the data transfer between the system and the device, whether the device is able to operate properly, etc.
In a legal knowledge retrieval device based on a knowledge graph, the construction of a database mainly comprises the following steps:
and (3) data collection: legal related data is collected, including legal provision, regulations, cases, decision documents, and the like. The data are acquired by means of web crawlers, data acquisition tools and the like.
Data cleaning: and cleaning and preprocessing the collected data, removing redundant information, formatting the data and the like to ensure the accuracy and consistency of the data.
Modeling data: according to the characteristics and requirements of the legal field, a data model of the database is designed. Modeling is performed by using a relational database, a graph database, and the like.
Determining the requirements and targets: characteristics of the legal field and problems to be solved are defined, and the target and the functional requirement of the database are determined.
Analysis data: analyzing the collected legal data, and knowing the structure and relation of the data, and the dependency relation and constraint condition between the data.
Design entity and properties: based on the results of the analysis, entities (Entities) in the database and Attributes (Attributes) of the entities are determined. An entity is a legal provision, a regulation, a case, a court, etc., and an attribute is a characteristic, description, or key information of the entity.
Design relationship: relationships (Relationships) between entities are determined, including one-to-one Relationships, one-to-many Relationships, and many-to-many Relationships. The relationship is a relationship between a rule and a legal provision, a relationship between a case and a legal provision, or the like.
Designing a main key and an external key: a primary Key (PRIMARY KEY) and a Foreign Key (Foreign Key) are determined for each entity. The primary key is an attribute that uniquely identifies an entity and the foreign key is an attribute that links relationships between entities.
Normalizing: according to the normal form theory of database design, the entity and the relation are normalized, redundant and repeated data are eliminated, and the performance of the database and the consistency of the data are improved.
Design index: and designing a proper index according to the query requirement and the data access mode so as to improve the retrieval speed and the query efficiency of the data.
Defining a constraint: corresponding constraint conditions, such as uniqueness constraint, non-null constraint, reference integrity constraint and the like, are defined according to the business requirements and the integrity requirements of the data.
Selecting a database type: and selecting a graph database according to actual requirements and data characteristics.
And according to the designed data model, the creation of the database and the establishment of the table structure are implemented, and the test and the tuning are carried out, so that the functions and the performances of the database are ensured to meet the requirements.
Data import: and importing the cleaned data into a database, and storing and organizing according to a data model designed in advance. The data importation may be accomplished using an importation tool of a database or by writing a script.
Indexing and optimizing: and indexing and optimizing the performance of the database to improve the retrieval speed and the query efficiency of the data. And an appropriate index can be created according to the query requirement, so that the query statement and the table structure are optimized, and the performance of the database is improved.
Data update and maintenance: legal data in the database is updated regularly, and timeliness and accuracy of the data are maintained. And meanwhile, the backup and maintenance of the database are carried out, and the safety and reliability of the data are ensured. It should be noted that the construction process of the database is an iterative process, and along with the continuous perfection and updating of the knowledge graph, the database also needs to be correspondingly updated and maintained so as to maintain the consistency and accuracy of the knowledge graph.
As shown in fig. 2-3, in the embodiment of the present invention, the map construction module further includes a knowledge fusion unit, a knowledge processing unit, a knowledge map construction unit and a knowledge storage unit, the knowledge fusion unit transmits data to the knowledge processing unit, the knowledge processing unit transmits data to the knowledge map construction unit, the knowledge map construction unit transmits data to the knowledge storage unit, the map update unit transmits data to the knowledge fusion unit, the knowledge fusion unit is used for integrating new knowledge, eliminating contradiction and ambiguity therein, for example, a certain object has various expressions, a certain description represents a plurality of objects, at this time, the integration processing is performed on the knowledge fusion data, the knowledge processing unit adds qualified parts into a database through quality evaluation, the knowledge map construction unit is used for building a relationship between knowledge to construct a map, and the knowledge storage unit is used for storing a knowledge question and transmitting the data.
The knowledge data acquired by the map construction module come from a legal knowledge base, the legal knowledge base receives a data request of the data acquisition module, relevant knowledge is acquired and is transmitted to the data acquisition module, the data acquisition module transmits acquired data to the data analysis module, the data analysis module transmits the data obtained through analysis to the data processing module, the data processing module transmits the data to the map construction module, the legal knowledge base is used for storing relevant knowledge of laws and is communicated with big data, a large number of and detailed legal knowledge resources are stored, the data acquisition module is used for acquiring required legal knowledge, the data analysis module is used for carrying out data analysis on the acquired data, analyzing the types and relevant relations of the data, and the data processing module is used for preprocessing the data and providing assistance for integration of subsequent data.
The visual display module displays the knowledge graph, mainly provides the relation between the entities, the related knowledge in the current field and the like for the user, for example, when searching the national fire, the knowledge related to the national fire is displayed, and meanwhile, the related knowledge is marked with the eye-catching mark, so that the user can acquire legal knowledge conveniently.
The knowledge graph construction unit is used for linking the knowledge with the keywords, and during data analysis, firstly judging the search content, judging which legal knowledge is needed for the content, if search distribution is carried out, firstly judging which legal knowledge relates to the respective problems, and meanwhile summarizing and classifying the knowledge, so that an intuitive knowledge acquisition way is provided for a user.
The visualization module is provided with a personalized management and typesetting optimization module for displaying the atlas, the interface is optimized through personalized management, the user can conveniently find and acquire information, meanwhile, in the aspect of typesetting, an administrator optimizes typesetting according to user feedback, if certain inquiry or knowledge acquisition links are not easy to find or display content typesetting is not visual enough, and the administrator can optimize according to reality.
When the user uses the map updating unit each time, the collected information is connected with the search word and transmitted to the knowledge fusion unit to reestablish and update the map, and the knowledge relationship added by the map updating unit for the map is more convenient for the user to acquire the knowledge along with the increase of the use times of the user.
The user uses the touch screen 2, touches the touch screen, the electric signal generated by touching is transmitted to the input analysis module through the information interface, meanwhile, when the user searches related knowledge, text input is carried out on the touch screen, the content to be searched is input, the input content is converted into data which can be identified by a computer through the information interface and is transmitted to the input analysis module, the input analysis module analyzes the data transmitted by the touch screen, text analysis is carried out on text information input by the user, data matching is carried out on the data, the search module searches the internal database, and the searched data is transmitted to the touch screen for display;
The system comprises a map construction module, a data acquisition module, a data analysis module, a data processing module and a data processing module, wherein knowledge data acquired by the map construction module are from a legal knowledge base, the legal knowledge base is used for storing relevant knowledge of laws and is communicated with big data, a large number of detailed legal knowledge resources are stored, the data acquisition module sends a data calling request to the legal knowledge database, the legal database invokes the relevant knowledge to transmit the data request to the data acquisition module after receiving the data request of the data acquisition module, the data acquisition module transmits acquired data to the data analysis module, the data analysis module classifies the data according to the data characteristics of the data, analyzes the relations among different legal knowledge, transmits the analyzed data to the data processing module, and the data processing module transmits the data to the map construction module, firstly, the wavelet analysis is adopted for denoising and compression of the big data, and the redundant repeated parts of the data and the information data with obvious text errors in the text information data transmitted by the legal knowledge base are removed, so that the accuracy of the big data is prevented from being influenced by noise; the multi-scale analysis using wavelet analysis can effectively overcome the block effect and mosquito noise, the wavelet basis function expression is Wherein a and b represent constants, the value range is [0, + ], then, the problem data in the big data are preprocessed, the problem data refer to data with obvious errors and repetition, the problem data are repaired by using an approximate repair method, and the integrity and accuracy return of the text information data of the database are ensured; in the application process of the approximate restoration method, no influence of other accidental factors must be ensured, and a calculation formula of the average value of the data in the adjacent time period is set
When the data feature extraction is carried out, a genetic algorithm is applied, firstly, the population scale is set as N in the first step, and the initial population isSimultaneously initializing a processing solution space, and providing an environment for feature selection of the initialization data; the second step is to calculate the individual fitness value of the t generation according to the moderate function, and record as/>The fitness function expression is as follows: wherein c represents the number of classified features themselves; /(I) The mean value vector of the subset in the class i is shown; /(I)The representation is an average vector of the feature set; n i represents the number of features in the i category; /(I)Representing a vector of j features in the i category; calculating individual fitness values to obtain each characteristic condition of the text information data of the database; thirdly, estimating the probability of the data characteristic, wherein the calculation formula is/>Wherein ω (x i) represents a weight value corresponding to the feature x i, and determining the proportion of each feature in the data text information; step four, selecting two individuals from the updated population according to the calculation result of the step three, carrying out cross recombination and mutation treatment on individual gene positions so as to obtain a new generation population, extracting the individuals corresponding to the maximum fitness value, and marking the individuals as/>The step extracts the text information data characteristics of the database with larger weight, and the extracted individual is the characteristic with higher importance degree of the current judgment; fifth step will/>Comparing the generated value with the fitness smell value, if the generated value is larger than or equal to a threshold value, terminating the iterative process, and outputting K features with the top ranking of the fitness to form an optimal feature subset; otherwise, turning to the second step; and repeating the iteration until the optimal data feature subset is selected.
The map construction module comprises a knowledge fusion unit, a knowledge processing unit, a knowledge map construction unit and a knowledge storage unit, wherein the knowledge fusion unit adopts a knowledge fusion method matched according to a dictionary and similarity, and divides the entities into two types of entity pairs and multiple groups after a relation extraction task, and the entity pairs in the multiple groups are identified to be required to additionally judge whether residual multiple information exists after the entity pairs pass through the entity pairs production task, so that the supernode pairs are completed, namely, the multiple definition relation residual is eliminated; firstly, sequentially inputting entities in a group to judge the type of the entities, if the input entity is of a legal type, carrying out matching search on the input entity by utilizing a legal alias dictionary, if the input entity is found, replacing the input entity according to a main name in the dictionary, and if the input entity is not found, entering similarity matching calculation; for entity types containing numerical information, standardization processing is needed, and if the entity types are the same as the updated entity expression, the entity types are regarded as the same entity; aiming at legal case field data, the alignment processing of the two types of entities can eliminate a larger proportion of residual information, so that the residual type entities adopt a quick and good-effect cosine similarity algorithm to match the similarity among the entities, the vectorization representation of the entity vocabulary level is completed by using a related tool, and the cosine similarity S between the entities and the entities of the same type in the atlas is sequentially calculated; the larger S is the stronger the cosine similarity of the two vectors, namely the closer the semantics of the two entities are, the entity which exceeds the set smell value and has the largest value is selected as a fusion object, otherwise, the entity which is failed to be matched is used as a new entity, and the processes are sequentially circulated to complete the alignment task of the entities in the group. Because the above processing may have a certain error, to ensure accuracy, it should be assisted to verify whether the entity that is not successfully matched needs entity alignment after the cycle is completed; if the entity in the entity pair completes the entity pair text task, a new triplet is obtained or linked to the existing node in the map to complete database updating. If the entity in the multi-element group is the entity, the super-node alignment is continuously finished, namely whether the entity set in the multi-element limiting relation under the super-node in the traversing map has the containing relation with the input multi-element group set or not, if the entity set does not have the containing relation, the entity set is stored in the knowledge map in the super-node form as a new multi-element limiting relation, otherwise, the super-set of the entity set and the input multi-element group is taken as the content stored in the map to update the database, and after knowledge fusion, the quality of the knowledge elements is improved, and the storage space is saved.
The knowledge processing unit is used for transmitting the data to the knowledge processing unit, the knowledge processing unit is used for transmitting the data to the knowledge map construction unit, the knowledge map construction unit is used for transmitting the data to the knowledge storage unit, the map updating unit is used for transmitting the data to the knowledge fusion unit, the knowledge fusion unit is used for integrating new knowledge, contradiction and ambiguity are eliminated, the knowledge processing unit is used for adding qualified parts into the database through quality evaluation on the knowledge fused data, the knowledge map construction unit is used for constructing the relationship of the knowledge so as to construct a map, and the knowledge storage unit is used for storing the knowledge problem shop and transmitting the data to the internal database for storage;
The visualization module is provided with a personalized management and typesetting optimization module for displaying the atlas, the interface is optimized through personalized management, the user can conveniently find and acquire information, meanwhile, in the aspect of typesetting, an administrator optimizes typesetting according to user feedback, if certain inquiry or knowledge acquisition links are not easy to find or display content typesetting is not visual enough, and the administrator can optimize according to reality.
Although embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made therein without departing from the principles and spirit of the invention, the scope of which is defined in the appended claims and their equivalents.
Claims (8)
1. Legal knowledge retrieval device based on knowledge graph is characterized by comprising:
A device main body (1) for supporting other components as the whole main body of the device;
The touch screen (2) is used for interaction between a user and a system connected with the device;
Wheels (3) for moving the device and for fixing the position;
The searching module is used for searching related legal knowledge by a user through touch control;
The map construction module is used for constructing the knowledge map before providing search and continuously updating the map in the process of search;
The internal database is used for storing related information of the established knowledge graph;
the knowledge reasoning module is used for supplementing and perfecting the newly established knowledge graph;
the visual display module is used for visually displaying the searched content and simultaneously visualizing the map for further searching;
The device comprises a device main body (1), wherein a touch screen (2) is arranged outside the device main body (1), wheels (3) are arranged outside the device main body (1), the touch screen obtains touch information after a user touches the device main body and transmits the information to a search module, the search module sends a data calling request to an internal database, the internal database responds to a signal after receiving the request signal, the internal database calls legal knowledge related to search content and transmits the legal knowledge to the touch screen for display, the internal database obtains and stores a constructed knowledge graph from a graph construction module, the graph construction module transmits the constructed preliminary graph to a knowledge reasoning module, the knowledge reasoning module transmits the supplemented knowledge graph to the graph construction module, and the graph construction module transmits data to a visual display module which transmits the data to the internal database.
2. The knowledge graph-based legal knowledge retrieval device according to claim 1, wherein the touch screen (2) transmits touch information to an information interface, the information interface transmits data to an input analysis module, the input analysis module transmits the analyzed data to a search module, the information interface application receives information transmitted by the touch screen and transmits the information to a system, the input analysis module is used for analyzing the data transmitted by the touch screen, performing text analysis on text information input by a user, and the search module acquires the text information transmitted by the input analysis and retrieves the information.
3. The knowledge-based legal knowledge retrieval device according to claim 1, wherein the graph construction module further comprises a knowledge fusion unit, a knowledge processing unit, a knowledge graph construction unit and a knowledge storage unit, wherein the knowledge fusion unit transmits data to the knowledge processing unit, the knowledge processing unit transmits data to the knowledge graph construction unit, the knowledge graph construction unit transmits data to the knowledge storage unit, the graph update unit transmits data to the knowledge fusion unit, the knowledge fusion unit is used for integrating new knowledge to eliminate contradiction and ambiguity, for example, a certain object has various expressions, a certain description represents a plurality of objects, at the moment, the integration processing is performed on the object, the knowledge processing unit adds qualified parts to a database through quality evaluation, the knowledge graph construction unit is used for building relations of knowledge to construct graphs, and the knowledge storage unit is used for storing knowledge questions and transmitting the data.
4. The knowledge graph-based legal knowledge retrieval device according to claim 1, wherein knowledge data acquired by the graph construction module is from a legal knowledge base, the legal knowledge base receives a data request of the data acquisition module to call related knowledge and transmit the data to the data acquisition module, the data acquisition module transmits acquired data to the data analysis module, the data analysis module transmits the data obtained through analysis to the data processing module, the data processing module transmits the data to the graph construction module, the legal knowledge base is used for storing related knowledge of laws and is communicated with big data, a large amount of detailed legal knowledge resources are stored, the data acquisition module is used for acquiring required legal knowledge, the data analysis module is used for carrying out data analysis on the acquired data, carrying out category classification and related relation analysis on the data, and the data processing module is used for preprocessing the data and providing assistance for integration of subsequent data.
5. The knowledge graph-based legal knowledge retrieval device according to claim 1, wherein the visual display module displays the knowledge graph, mainly provides the user with the relation between the entities and the knowledge related to the current field, for example, when searching for the national code, the knowledge related to the national code is displayed, and meanwhile, the knowledge related to the national code is marked for highlighting, so that the user can acquire legal knowledge conveniently.
6. The knowledge graph-based legal knowledge retrieval device according to claim 3, wherein the knowledge graph construction unit links knowledge with keywords, and when data analysis is performed, firstly judges search content, judges which legal knowledge is needed for the content, if search distribution is performed, firstly judges which legal knowledge involves respective problems, and simultaneously performs summary classification on the knowledge, so that an intuitive knowledge acquisition path is provided for users.
7. The knowledge graph-based legal knowledge retrieval device according to claim 5, wherein the visualization module is provided with a personalized management and typesetting optimization module for displaying graphs, and the interface is optimized through personalized management, so that a user can conveniently find and acquire information, and meanwhile, an administrator optimizes typesetting according to user feedback in typesetting, for example, certain inquiry or knowledge acquisition links are not easy to find or display content typesetting is not intuitive enough, and the administrator can optimize according to actual practice.
8. The legal knowledge retrieval device based on the knowledge graph according to claim 3, wherein the graph updating unit is used for collecting the information related to the search word and transmitting the information to the knowledge fusion unit to reestablish and update the graph when the user uses the graph, and the graph updating unit is used for adding the knowledge relationship to the graph to facilitate the user to acquire the knowledge along with the increase of the use times of the user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410067417.3A CN117908637A (en) | 2024-01-17 | 2024-01-17 | Legal knowledge retrieval device based on knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202410067417.3A CN117908637A (en) | 2024-01-17 | 2024-01-17 | Legal knowledge retrieval device based on knowledge graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117908637A true CN117908637A (en) | 2024-04-19 |
Family
ID=90691939
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202410067417.3A Pending CN117908637A (en) | 2024-01-17 | 2024-01-17 | Legal knowledge retrieval device based on knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117908637A (en) |
-
2024
- 2024-01-17 CN CN202410067417.3A patent/CN117908637A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108804521B (en) | Knowledge graph-based question-answering method and agricultural encyclopedia question-answering system | |
| CN111428054A (en) | Construction and storage method of knowledge graph in network space security field | |
| KR101508260B1 (en) | Summary generation apparatus and method reflecting document feature | |
| CN113779272B (en) | Knowledge graph-based data processing method, device, equipment and storage medium | |
| CN109408578B (en) | Monitoring data fusion method for heterogeneous environment | |
| CN110765277B (en) | Knowledge-graph-based mobile terminal online equipment fault diagnosis method | |
| CN104699767B (en) | A kind of extensive Ontology Mapping Method towards Chinese language | |
| Movshovitz-Attias et al. | Kb-lda: Jointly learning a knowledge base of hierarchy, relations, and facts | |
| CN113254507B (en) | Intelligent construction and inventory method for data asset directory | |
| CA2906767A1 (en) | Non-deterministic disambiguation and matching of business locale data | |
| CN112948547A (en) | Logging knowledge graph construction query method, device, equipment and storage medium | |
| CN114332519A (en) | Image description generation method based on external triple and abstract relation | |
| CN110737779A (en) | Knowledge graph construction method and device, storage medium and electronic equipment | |
| Partyka et al. | Enhanced geographically typed semantic schema matching | |
| CN113610626A (en) | Bank credit risk identification knowledge graph construction method and device, computer equipment and computer readable storage medium | |
| CN113377739A (en) | Knowledge graph application method, knowledge graph application platform, electronic equipment and storage medium | |
| CN109977235B (en) | Method and device for determining trigger word | |
| Zhu et al. | A data cleaning method for heterogeneous attribute fusion and record linkage | |
| CN112015792A (en) | Material duplicate code analysis method and device and computer storage medium | |
| CN114443783B (en) | Supply chain data analysis and enhancement processing method and device | |
| WO2021250950A1 (en) | Method, system, and device for evaluating performance of document search | |
| CN117908637A (en) | Legal knowledge retrieval device based on knowledge graph | |
| CN112613318B (en) | Entity name normalization system, method thereof and computer readable medium | |
| Lehmberg | Web table integration and profiling for knowledge base augmentation | |
| Wang et al. | A Markov logic network method for reconstructing association rule-mining tasks in library book recommendation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication |