[0056] The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. The specific embodiments described herein are only used to explain the present invention, and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.
[0057] The invention provides a system for assisting the generation of domain knowledge map, including a basic library module, a knowledge map auxiliary construction module, a submitted data integration processing module and a knowledge map output module. The basic library module includes a knowledge construction model library, a knowledge construction algorithm library, The domain knowledge base and the knowledge graph output symbol base, the knowledge building model base, the knowledge building algorithm base and the domain knowledge base are electrically connected with the knowledge graph auxiliary building module, the knowledge graph auxiliary building module is electrically connected with the submitted data integration processing module, and the submitted data Both the integrated processing module and the knowledge graph output symbol library are electrically connected with the knowledge graph output module;
[0058] The knowledge graph auxiliary construction module includes knowledge topic setting, keyword extraction, knowledge construction model recommendation and selection, and knowledge graph auxiliary construction; the submitted data integration processing module includes the basic identification of knowledge entities, the static space domain of knowledge entities, and the dynamic knowledge of knowledge entities. The time domain, the research domain of multiple knowledge entities, and the contact system of the knowledge entities, the submitted data integration processing module is used for the data integration processing of all node contents and construction models submitted by users for a given knowledge topic; the knowledge graph output module uses In order to integrate the data processed by the submitted data integration processing module, it is output in the form of knowledge graph for users to query, learn and use. In the final output domain knowledge graph, the knowledge graph output module displays various relationships in text or through the The symbols in the knowledge graph output symbol library are converted to symbols to improve the visualization performance of the graph.
[0059] The knowledge building model library includes conceptual model, attribute model, characteristic model, quantitative model, intuitive model, emotional model, structural model, functional model, upper model, lower model, parallel model, overall model, local model, common model, positive and negative models , Analytical Model, Comprehensive Model, Abstract Model, Generalization Model, Induction Model, Transformation Model, Similarity Model, Difference Model, Similarity Model, Class Shift Model, Analogy Model, Process Model, Result Model, Cause Model, Background Model, Condition model and measure model, knowledge construction model library is used to analyze and construct a certain knowledge entity in the field.
[0060] The knowledge construction algorithm library includes conceptual algorithm, attribute algorithm, feature algorithm, quantitative algorithm, intuitive algorithm, emotional algorithm, structural algorithm, functional algorithm, upper-level algorithm, lower-level algorithm, parallel algorithm, overall algorithm, local algorithm, common algorithm, positive and negative algorithm , analysis algorithm, synthesis algorithm, abstract algorithm, generalization algorithm, induction algorithm, deductive algorithm, associative algorithm, imaginary algorithm, transformation algorithm, algorithm for seeking common ground, algorithm for seeking difference, similarity algorithm, class shift algorithm, analogy algorithm, process algorithm, The result algorithm, the cause algorithm, the background algorithm, the condition algorithm and the measure algorithm, and the knowledge construction algorithm base is used to extract the relevant content of a knowledge entity in the domain from the domain knowledge base.
[0061] The domain knowledge base is a large domain knowledge base established by the knowledge construction model in the knowledge construction model base, artificial addition or network collection and processing methods, and provides users with a reference in the process of creating domain knowledge graphs.
[0062] The knowledge graph output symbol library includes central topic, topic content, descriptive content, development process/stage, item list, content list, connecting line, arrow line, concept, attribute, connection, feature, internal feature, external feature, quantity , intuition, emotion, structure, composition, essence, status, function, superior, inferior, juxtaposition, whole, part, common, positive and negative, analysis, synthesis, abstraction, generalization, induction, deduction, association, imagination, transformation, seeking Similarity, difference, similarity, class shift, analogy, process, result, reason, background, conditions and measures, the knowledge graph output symbol library is used to symbolize various relationships in the domain knowledge graph.
[0063] The knowledge graph auxiliary building module is used for the user's specific domain knowledge graph generation operation, which includes the following steps:
[0064] A1. Knowledge topic setting, which is used for users to input knowledge topics that need to generate domain knowledge graphs, and use them as first-level nodes;
[0065] A2. Keyword extraction, which is used for keyword extraction of the knowledge topic of the domain knowledge graph input by the user;
[0066] A3. Knowledge construction model recommendation and selection, which is used by the system to recommend or select the knowledge construction model by the user through the analysis of the acquired knowledge subject keywords;
[0067] A4. Secondary node generation is used to systematically combine the acquired knowledge topic keywords and the knowledge construction model finally determined by the user, and obtain matching knowledge topic keywords from the domain knowledge base through the knowledge construction algorithm corresponding to the knowledge construction model. , and the subordinate knowledge topic corresponding to the knowledge construction model is used as the reference for the user to construct the subordinate knowledge topic, and the user determines the specific content of this node according to the specific reference content;
[0068] A5. According to the specific content determined by the secondary node, repeat A2 to A4 to generate the third-level node, and repeat this cycle until it can no longer continue, and the generation of this group of nodes is completed;
[0069] A6. Repeat A3 to A4 to generate other secondary nodes different from the previous group, and generate each group of nodes according to A5 until no new selection can be made, and the domain knowledge graph is assisted to complete the construction.
[0070] Among them, when generating second-level nodes, the same knowledge construction model can be used to analyze the first-level knowledge topics from different angles and aspects, and each construction model is not necessarily used, and the specific use depends on the attributes of the nodes. OK, the same is true for lower-level nodes; when generating second-level nodes, the system's recommendation for building models will be grouped as follows:
[0071] A. The construction of internal feature content of knowledge entities, including conceptual model, attribute model, structural model, local model, and subordinate model;
[0072] B. Construction of external feature content of knowledge entities, including feature model, intuitive model, quantitative model, and emotional model;
[0073] C. Knowledge entity dynamic development content construction, including process model;
[0074] D. Rational knowledge content construction of knowledge entities, including comprehensive models, analytical models, abstract models, generalization models, inductive models, and deductive models;
[0075] E. Content construction of the existence status of knowledge entities, including functional model, result model, cause model, background model, condition model, and measure model;
[0076] F. Contrastive content construction of knowledge entities, including difference model, common model, similarity model, analogy model, and class-shift model;
[0077] G. Peripheral expansion content construction of knowledge entities, including upper model, parallel model, overall model, common model, positive and negative model, transformation model, associative model, and imagination model.
[0078] The basic identification of knowledge entities mainly includes the concept and attributes of entities, and the study of entities from the perspective of systems theory; such as figure 2As shown, attributes are the core content of knowledge entities, including the nature of entities and the relationship between entities and other entities; entities can be divided into individual entities and type entities according to their composition, among which, type entities are reflected by concepts; For the type entity, it is necessary to carry out classification research; for the individual entity, it is necessary to decompose it; from the perspective of system theory, any entity is a system, and the research focus of the system lies in the study of its structure and function; the function of the system is essentially It is its status or role in the environment. For knowledge entities, it is the relationship between the entity and other entities; the structure of the system includes its constituent elements and the relationship between elements and between elements and the system. For knowledge entities, it is the relationship between the entity and other entities. the nature of the entity.
[0079] The static space domain of knowledge entities is studied from the attributes of the entity itself and the structure and function based on system theory; such as image 3 As shown, starting from the attributes of knowledge entities, to study its characteristics, from two aspects: internal and external: external characteristics depend on internal characteristics to a certain extent, which can be intuitively perceived; Carry out quantitative (quantitative) analysis; on the basis of quantitative analysis and intuitive perception of the nature and characteristics of knowledge entities, analyze people's emotional feedback on it, and further form corresponding ideas. Starting from the attributes of a knowledge entity, to study its relationship with other entities, from two aspects: human beings and other entities: for humans, focus on analyzing its use; for other entities, it needs to analyze its function and meaning . From the perspective of systems theory, its function is studied in combination with its use to humans and its role and meaning in other entities. From the perspective of system theory, through the analysis of the structure, clarify its composition, essence, status and influence on internal characteristics; on the basis of the analysis, abstract and obtain its essential attributes; and through generalization, obtain its concepts, principles, rules, etc.
[0080] The dynamic time domain of knowledge entities is the study of entities from the perspective of time; such as Figure 4 From the perspective of time, a certain state of development of any entity is produced by a certain process under a certain cause; at the same time, its existence has a certain position in the external environment where it is located and has a certain impact on other The generation or occurrence of the entity has an impact; among the many factors that cause the generation or occurrence of the knowledge entity, the focus will be on the large external environmental factors (background) that generate or occur, and the favorable factors (conditions) that promote it; In addition, in order to ensure that it can develop in the desired state in the future, it needs to be studied for measures.
[0081] The research domain of multiple knowledge entities is mainly carried out from three aspects: the generation of multiple entities, the research of multiple entities in the same system, and the research of multiple entities in different systems; such as Figure 5 As shown, for known knowledge entities, other entities can be deduced through transformation, association, and imagination. In the same system, through the synthesis of multiple entities, the overall characteristics of the entire system can be obtained; through the induction of multiple entities, the general principles and conclusions of the system can be drawn; on the basis of induction, further general principles of the system can be obtained. , the conclusion can be deduced, and other individual characteristics of the entity can be deduced. In different systems, through comparison, their identical (similar), similar and different (different) features can be obtained; at the same time, based on their identical and similar features, new entities can be formed through class shift; through analogy , they can be deduced to be identical or similar in other features.
[0082] The connection system of knowledge entities is mainly carried out from the structure of system theory and the correlation of various elements, such as Image 6 As shown, for any entity, the entity with which it is associated either belongs within it or exists outside of it. Whether it is internal or external, it can be attributed to the system, that is, the entity itself is either a system or a part of a system. Therefore, for the connection between entities, we can abstract it into two basic connection methods from the viewpoint of system theory and the method of structural analysis: the connection between the whole and the part, and the connection between the part and the part. Starting from the structure of the system, to study the connection between various entities, it needs to be carried out from the perspectives of space, time and nature:
[0083] From the perspective of space, the connection within the system is mainly manifested in the connection between the elements (parts) that make up the system and the connection with the system (the whole). Since the system is divided into types and individuals, the connections between them are mainly divided into four categories: upper and lower connections, overall and partial connections, parallel connections, and common connections.
[0084] From the perspective of time, the connection within the system is mainly manifested in the connection between different stages (parts) of system development and the whole development process (whole) of the system. Due to the sequence and irreversibility of the developmental stages of general systems, the connections between them are mainly the connection between the whole and the part, and the connection between the part and the part that is reflected in the causal connection.
[0085] From the perspective of nature, the connection within the system is mainly manifested in the connection between the different properties (parts) of the system and the overall nature of the system. Starting from the correlation between various elements in the system, to study the connection between various entities, it is mainly manifested in the following forms:
[0086] Contradiction, showing the contradictory relationship between the elements, which is reflected in the partial and partial connection of positive and negative links;
[0087] Opposition, showing the common or juxtaposed relationship between elements, which is reflected in the connection between parts;
[0088] Inclusiveness, showing the relationship between the upper and lower levels or the whole and the part among the elements, which is reflected in the connection between the whole and the part;
[0089] Identity, showing the identical relationship between elements, which is reflected in the part and part of the equivalence connection;
[0090] Intersection, which shows the same relationship of elements at a certain angle, is reflected in the part and part of the similar connection.
[0091] Based on the above-mentioned system for assisting in generating a domain knowledge graph, the present invention also provides a method for assisting in generating a domain knowledge graph, comprising the following steps:
[0092] S1. The user first uses the knowledge graph to assist the knowledge topic setting operation in the building module, and uses pinyin, handwriting, voice input, etc. to input the knowledge topic that needs to generate a domain knowledge graph, as a first-level node;
[0093] S2. After the user confirms the input, the keyword extraction operation will perform keyword extraction on the knowledge topic of the domain knowledge graph set by the user;
[0094] S3. The system recommends or the user chooses the knowledge to build a model by analyzing the acquired knowledge subject keywords;
[0095] S4, the user determines the knowledge construction model;
[0096] S5. The system combines the acquired knowledge topic keywords and the knowledge construction model finally determined by the user, and obtains the matching knowledge topic keywords from the domain knowledge base through the knowledge construction algorithm corresponding to the knowledge construction model, and matches the knowledge construction model. The corresponding subordinate knowledge topics are used as references for users to construct subordinate knowledge topics;
[0097] S6. The user determines the specific content of the secondary node according to the reference content given by the system;
[0098] S7. After the second-level node is determined, repeat S2 to S6 to generate the third-level node, and repeat this cycle until it can no longer continue, and the generation of this group of nodes is completed;
[0099] S8. Repeat S3 to S7 to generate other secondary nodes different from the previous group, and generate each group of nodes according to S7 until no new selection can be made, and the domain knowledge graph is assisted to complete the construction;
[0100] S9. The submitted data integration processing module processes the data submitted by the user;
[0101] S10, the domain knowledge graph output module outputs the final knowledge graph.
[0102] Finally, it should be noted that the above descriptions are only preferred embodiments of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, for those skilled in the art, the The technical solutions described in the foregoing embodiments can be modified, or some technical features thereof can be equivalently replaced, and any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included. within the protection scope of the present invention.