CN117436531A - Question answering system and method based on rice pest knowledge graph - Google Patents
Question answering system and method based on rice pest knowledge graph Download PDFInfo
- Publication number
- CN117436531A CN117436531A CN202311763394.1A CN202311763394A CN117436531A CN 117436531 A CN117436531 A CN 117436531A CN 202311763394 A CN202311763394 A CN 202311763394A CN 117436531 A CN117436531 A CN 117436531A
- Authority
- CN
- China
- Prior art keywords
- entity
- question
- unit
- knowledge graph
- relation
- 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
- 241000607479 Yersinia pestis Species 0.000 title claims abstract description 131
- 235000007164 Oryza sativa Nutrition 0.000 title claims abstract description 71
- 235000009566 rice Nutrition 0.000 title claims abstract description 71
- 238000000034 method Methods 0.000 title claims abstract description 49
- 240000007594 Oryza sativa Species 0.000 title 1
- 241000209094 Oryza Species 0.000 claims abstract description 99
- 201000010099 disease Diseases 0.000 claims abstract description 82
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 claims abstract description 82
- 238000004458 analytical method Methods 0.000 claims abstract description 32
- 238000000605 extraction Methods 0.000 claims abstract description 26
- 241000238631 Hexapoda Species 0.000 claims description 59
- 238000012545 processing Methods 0.000 claims description 23
- 238000012795 verification Methods 0.000 claims description 21
- 238000012800 visualization Methods 0.000 claims description 17
- 238000007781 pre-processing Methods 0.000 claims description 12
- 241000196324 Embryophyta Species 0.000 claims description 10
- 238000012549 training Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 6
- 238000001228 spectrum Methods 0.000 claims description 4
- 238000005516 engineering process Methods 0.000 abstract description 7
- 238000010586 diagram Methods 0.000 description 21
- 230000008569 process Effects 0.000 description 12
- 238000004590 computer program Methods 0.000 description 11
- 230000006870 function Effects 0.000 description 9
- 238000013135 deep learning Methods 0.000 description 8
- 230000005540 biological transmission Effects 0.000 description 7
- 238000004422 calculation algorithm Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000013528 artificial neural network Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 235000013339 cereals Nutrition 0.000 description 3
- 238000011161 development Methods 0.000 description 3
- 230000018109 developmental process Effects 0.000 description 3
- 238000003058 natural language processing Methods 0.000 description 3
- 238000012360 testing method Methods 0.000 description 3
- 238000004891 communication Methods 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000004519 manufacturing process Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 239000013598 vector Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 208000036752 Schizophrenia, paranoid type Diseases 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 230000006835 compression Effects 0.000 description 1
- 238000007906 compression Methods 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000008676 import Effects 0.000 description 1
- 230000003993 interaction Effects 0.000 description 1
- 238000010380 label transfer Methods 0.000 description 1
- 238000002372 labelling Methods 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 208000002851 paranoid schizophrenia Diseases 0.000 description 1
- 239000000575 pesticide Substances 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000000547 structure data Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/04—Inference or reasoning models
- G06N5/041—Abduction
-
- 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
- G06F16/33—Querying
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/211—Syntactic parsing, e.g. based on context-free grammar [CFG] or unification grammars
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N5/00—Computing arrangements using knowledge-based models
- G06N5/02—Knowledge representation; Symbolic representation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q50/00—Information and communication technology [ICT] specially adapted for implementation of business processes of specific business sectors, e.g. utilities or tourism
- G06Q50/02—Agriculture; Fishing; Forestry; Mining
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Business, Economics & Management (AREA)
- Life Sciences & Earth Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Economics (AREA)
- Mining & Mineral Resources (AREA)
- Marine Sciences & Fisheries (AREA)
- Animal Husbandry (AREA)
- Agronomy & Crop Science (AREA)
- Human Resources & Organizations (AREA)
- Marketing (AREA)
- Primary Health Care (AREA)
- Strategic Management (AREA)
- Tourism & Hospitality (AREA)
- General Business, Economics & Management (AREA)
- Animal Behavior & Ethology (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application relates to a question answering system and a question answering method based on a rice pest knowledge graph, wherein the question answering system based on the rice pest knowledge graph comprises the following steps: a disease and pest relation question-answering module; the pest and disease relation question-answering module comprises a first interface unit, an information extraction unit and a retrieval unit; the first interface unit is used for acquiring a current problem input by a user and outputting a reply corresponding to the current problem; the information extraction unit is used for extracting a main entity in the current problem and extracting a target relation in the current problem through dependency syntax analysis; the searching unit is used for searching a target guest entity with the target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity. By the method and the device, the problem that the answers of the rice pest question-answering system in the related technology lack of interpretability is solved.
Description
Technical Field
The application relates to the field of question answering systems, in particular to a question answering system and method based on a rice pest knowledge graph.
Background
The question-answering system is very important research content in the fields of deep learning, artificial intelligence and the like, and particularly relates to a computer which is used for intelligently analyzing information such as voice, text, video and the like input by a user, searching based on a built knowledge base and finally outputting answers required by the user according to search results. The development of the question-answering system is roughly divided into two stages, namely a method based on context matching and a modern NLP (natural language processing) based method. In early stage, intelligent question and answer is realized by manually constructing a question library and an answer library and manually writing answer rules. For example, the university of Stanford developed a question-and-answer system for simulating paranoid schizophrenia manifestations, PARRY, wilensky et al developed a question-and-answer system called UC (UNIX Consultant). However, this technique requires excessive manual intervention, and only answers questions within the rule, so that the generalization capability is not strong enough, and the accuracy of question identification and answer output needs to be improved. After deep learning is raised, various models and algorithms are applied to each stage of the question-answering system, so that the accuracy and the answer speed of the system are improved. The NLP-based method mainly comprises the steps of extracting characteristics of a text through deep learning, and then completing the realization of a question-answering system by combining the characteristics with various pre-training models, neural networks, knowledge patterns and other technologies.
For the field of rice diseases and insect pests, a corresponding question-answering system is not available at present. Although the question-answering system based on deep learning can be directly applied to the field of rice diseases and insect pests. However, algorithms and models in deep learning have a "black box effect" that can result in a computation process that is not transparent. Thus, such methods are data-driven, and as a result lack of interpretability, the answers of the rice pest question-answering system lack of interpretability and reliability.
For the problem of lack of interpretability of answers of a rice pest question-answering system in the related art, no effective solution is proposed at present.
Disclosure of Invention
In the embodiment, a question answering system, a question answering method and electronic equipment based on a rice pest knowledge graph are provided, so that the problem that the answer of the question answering system for rice pests lacks interpretability in the related technology is solved.
In a first aspect, the invention provides a question-answering system based on a knowledge graph of rice diseases and insect pests, the system comprising: a disease and pest relation question-answering module;
the pest and disease relation question-answering module comprises a first interface unit, an information extraction unit and a retrieval unit;
the first interface unit is used for acquiring a current problem input by a user and outputting a reply corresponding to the current problem;
The information extraction unit is used for extracting a main entity in the current problem and extracting a target relation in the current problem through dependency syntax analysis;
the searching unit is used for searching a target guest entity with the target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
In some of these embodiments, the knowledge-graph is stored in a Neo4j graph database.
In some of these embodiments, the system further comprises: a data processing module;
the data processing module comprises a data preprocessing unit and a knowledge graph construction unit;
the data preprocessing unit is used for preprocessing the rice plant diseases and insect pests data;
the knowledge graph construction unit is used for analyzing the preprocessed rice plant diseases and insect pests data based on a pre-training language model, extracting entities and relations in the preprocessed rice plant diseases and insect pests data, and constructing the knowledge graph based on extraction results.
In some of these embodiments, the system further comprises: logging in a registration module;
the login registration module comprises a second interface unit, a data specification verification unit and a hash processing unit;
The second interface unit is used for acquiring login information or registration information input by a user;
the data specification checking unit is used for detecting whether the login information or the registration information accords with a specification;
the hash processing unit is used for carrying out hash encryption on the registration information when the registration information accords with the specification; and the hash verification is used for carrying out hash verification on the login information when the login information accords with the specification.
In some of these embodiments, the system further comprises: a named entity recognition module;
the named entity identification module comprises a third interface unit, a remote call model unit and a key pair verification unit;
the third interface unit is used for acquiring a text to be identified input by a user and outputting an entity identification result of the text to be identified;
the remote calling unit is used for establishing remote connection with a server, sending a text to be identified, which is input by a user, to the server, calling a named entity identification model in the server to identify the entity of the text to be identified, and acquiring an entity identification result from the server;
the key pair verification unit is used for guaranteeing the security of remote connection.
In some of these embodiments, the system further comprises: a plant diseases and insect pests relation retrieval module;
the plant diseases and insect pests relation retrieval module comprises a fourth interface unit and an entity relation matching unit;
the fourth interface unit is used for acquiring an entity to be queried input by a user, providing an entity selection list for the user, acquiring an entity selection instruction of the user for the entity selection list, determining the entity to be queried according to the entity selection instruction, and outputting a relation and an entity corresponding to the entity to be queried;
the entity relation matching unit is used for matching the relation and the entity corresponding to the entity to be queried in the knowledge graph.
In some of these embodiments, the system further comprises: a rice plant disease and insect pest visualization module;
the rice plant diseases and insect pests visualization module comprises a fifth interface unit and a visualization unit;
the fifth interface unit is used for acquiring a zoom instruction input by a user;
the visualization unit is used for displaying the knowledge graph to a user and scaling the knowledge graph according to the scaling instruction.
In some of these embodiments, the named entity recognition model includes a MA-RBC model.
In some of these embodiments, the extracting the master entity in the current question extracts the target relationship in the current question through dependency syntax analysis, including:
classifying the current problem to obtain the type of the current problem, wherein the type comprises a single relation problem or a complex relation problem;
identifying a master entity in the current question;
and extracting the target relation in the current problem according to the type of the current problem through dependency syntax analysis.
In a second aspect, the invention provides a question-answering method based on a knowledge graph of rice diseases and insect pests, which comprises the following steps:
acquiring a current problem input by a user;
extracting a main entity in the current problem, and extracting a target relation in the current problem through dependency syntax analysis;
and searching a target guest entity with the target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
In a third aspect, the invention provides an electronic device, which comprises a memory, a processor and a computer program stored in the memory and capable of running on the processor, wherein the processor realizes the question-answering method based on the rice pest knowledge graph in the second aspect when executing the computer program.
In a fourth aspect, the present invention provides a storage medium having stored thereon a computer program which, when executed by a processor, implements the question answering method based on the knowledge graph of rice plant diseases and insect pests according to the second aspect.
Compared with the related art, the question-answering system based on the rice pest knowledge graph firstly acquires the current question input by the user, carries out dependency syntactic analysis on the current question input by the user through the information extraction unit and the retrieval unit, carries out reconstruction semantic information, reduces technical complexity, carries out retrieval and reasoning on the knowledge graph based on knowledge driving, and enables the answer to have interpretability, and the algorithm and the model of deep learning in the prior art have 'black box effect', so that the transparency of the calculation process is not high, and the method is based on data driving and results lack of interpretability. Therefore, compared with the prior art, the question-answering system based on the rice pest knowledge graph has stronger interpretation and reliability. Solves the problem that the answers of the question answering system aiming at the rice diseases and insect pests in the related technology lack the interpretability.
The details of one or more embodiments of the application are set forth in the accompanying drawings and the description below to provide a more thorough understanding of the other features, objects, and advantages of the application.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiments of the application and together with the description serve to explain the application and do not constitute an undue limitation to the application. In the drawings:
FIG. 1 is a block diagram of a terminal hardware structure for executing a question-answering method based on a knowledge graph of rice plant diseases and insect pests provided by the invention;
FIG. 2 is a block diagram of a question-answering system based on knowledge graphs of rice diseases and insect pests of the present invention;
FIG. 3 is an algorithm chart of questions and answers based on the knowledge graph of rice diseases and insect pests;
FIG. 4 is a block diagram of a data processing module of the question-answering system based on the knowledge graph of rice plant diseases and insect pests of the present invention;
FIG. 5 is a block diagram of a login and registration module of the question-answering system based on the knowledge graph of rice diseases and insect pests;
FIG. 6 is a flow chart of a registration page of a question-answering system based on a rice pest knowledge graph of the invention;
FIG. 7 is a block diagram of a named entity recognition module of the question-answering system based on the knowledge graph of rice diseases and insect pests;
FIG. 8 is a flow chart of a named entity recognition page of the question-answering system based on the knowledge graph of rice diseases and insect pests;
FIG. 9 is a block diagram of a pest and disease relationship retrieval module of the question-answering system based on the rice pest and disease knowledge graph of the present invention;
FIG. 10 is a flow chart of a pest and disease relationship search page of the question-answering system based on the rice pest and disease knowledge graph of the present invention;
FIG. 11 is a block diagram of a rice pest visualization module of the question-answering system based on a rice pest knowledge graph of the present invention;
FIG. 12 is a flow chart of a rice pest visualization page of the question-answering system based on the rice pest knowledge graph of the present invention;
fig. 13 is a block diagram of a question-answering system based on knowledge graphs of rice plant diseases and insect pests according to the present invention.
Detailed Description
For a clearer understanding of the objects, technical solutions and advantages of the present application, the present application is described and illustrated below with reference to the accompanying drawings and examples.
Unless defined otherwise, technical or scientific terms used herein shall have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. The terms "a," "an," "the," "these," and the like in this application are not intended to be limiting in number, but rather are singular or plural. The terms "comprising," "including," "having," and any variations thereof, as used in the present application, are intended to cover a non-exclusive inclusion; for example, a process, method, and system, article, or apparatus that comprises a list of steps or modules (units) is not limited to the list of steps or modules (units), but may include other steps or modules (units) not listed or inherent to such process, method, article, or apparatus. The terms "connected," "coupled," and the like in this application are not limited to physical or mechanical connections, but may include electrical connections, whether direct or indirect. Reference to "a plurality" in this application means two or more. "and/or" describes an association relationship of an association object, meaning that there may be three relationships, e.g., "a and/or B" may mean: a exists alone, A and B exist together, and B exists alone. Typically, the character "/" indicates that the associated object is an "or" relationship. The terms "first," "second," "third," and the like, as referred to in this application, merely distinguish similar objects and do not represent a particular ordering of objects.
The method embodiments provided in the present invention may be performed in a terminal, a computer or similar computing device. For example, the method runs on a terminal, and fig. 1 is a block diagram of a terminal hardware structure for executing a question-answering method based on a rice pest knowledge graph provided by the invention. As shown in fig. 1, the terminal may include one or more (only one is shown in fig. 1) processors 120 and a memory 140 for storing data, wherein the processors 120 may include, but are not limited to, a processing device such as a microprocessor MCU or a programmable logic device FPGA. The terminal may further include a transmission device 160 for a communication function and an input-output device 180. It will be appreciated by those skilled in the art that the structure shown in fig. 1 is merely illustrative and is not intended to limit the structure of the terminal. For example, the terminal may also include more or fewer components than shown in fig. 1, or have a different configuration than shown in fig. 1.
The memory 140 may be used to store a computer program, for example, a software program of application software and a module, such as a computer program corresponding to a question-answering method based on a knowledge graph of rice plant diseases and insect pests in the present invention, and the processor 120 performs various functional applications and data processing, i.e., implements the above-mentioned method, by running the computer program stored in the memory 140. Memory 140 may include high-speed random access memory, and may also include non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some examples, memory 140 may further include memory located remotely from processor 120, which may be connected to the terminal via a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The transmission device 160 is used to receive or transmit data via a network. The network includes a wireless network provided by a communication provider of the terminal. In one example, the transmission device 160 includes a network adapter (Network Interface Controller, simply referred to as NIC) that may be connected to other network devices via a base station to communicate with the internet. In one example, the transmission device 160 may be a Radio Frequency (RF) module for communicating with the internet wirelessly.
In the invention, a question-answering system based on rice pest knowledge graph is provided, fig. 2 is a structural diagram of the question-answering system based on rice pest knowledge graph of the invention, as shown in fig. 2, the system comprises: a disease and pest relation question-answering module; the pest and disease relation question-answering module includes a first interface unit 210, an information extraction unit 220, and a retrieval unit 230; the first interface unit 210 is configured to obtain a current question input by a user, and output a reply corresponding to the current question; the information extraction unit 220 is used for extracting a main entity in the current problem and extracting a target relation in the current problem through dependency syntax analysis; the retrieving unit 230 is configured to retrieve a target guest entity having a target relationship with a host entity from a pre-constructed knowledge graph, and generate a reply of the current problem according to the target guest entity.
Specifically, the question-answering system based on the rice pest knowledge graph can be constructed based on a flash framework, wherein flash is a lightweight web development framework written based on a Werkzeug toolbox, and background task processing can be performed by using Python language. The WSGI toolbox adopts Werkzeug, and the template engine uses Jinja2. The front end may display the question and answer results at specific locations on the page using the jQuery framework. The jQuery framework is utilized to simplify DOM operation, and the function of dynamically updating text content when a user interacts with a page is realized.
For the first interface unit, a pest and disease relation question and answer page is provided for the user, and the page at least comprises a question input area and a result output area. For example, the pest relationship question-answering page may provide a text entry box for entering questions, with the results being displayed in the underlying blank area. After a user inputs a question, a question and answer result is displayed at the lower side, a relation diagram corresponding to a main entity and relation of the question is displayed at the left side, and detailed information of the entity is displayed at the right side. If the answer to the question is not found in the graph database, a prompt is given below.
And for the information extraction unit, after the first interface unit acquires the current problem input by the user, the information extraction unit extracts the information of the current problem. The information extraction unit extracts target relationships related to the master entity in the current problem mainly through dependency syntax analysis. The dependency syntactic analysis is a syntactic structure analysis technology, and by analyzing sentences, the sentences are decomposed into a plurality of components, and the components are modeled in a mode of describing a text grammar structure so as to better understand the syntactic structure of a sentence. Therefore, in the question-answering system, the dependency syntax analysis can rapidly and accurately extract the target relation related to the host entity from the current problem, thereby improving the retrieval accuracy of the guest entity. Therefore, the application of the dependency syntax analysis in the question-answering system is helpful to improve the answer efficiency and accuracy of the question-answering system, so that a user can acquire the required information faster and more conveniently.
For the retrieval unit, after the information extraction unit extracts the main entity of the current problem and the corresponding target relationship thereof, the retrieval unit retrieves the target guest entity having the target relationship with the main entity from the pre-constructed knowledge graph, and replies to the current problem. Illustratively, the jQuery framework will display answer text in a specified area. The background also returns corresponding JSON data, picture data and detailed information. The picture data is composed of pest pictures and is obtained in the data collection stage. And converting the JSON data into a corresponding relation diagram by using the ECharts framework. The detailed information mainly includes the name, pest location, place, etc. Preferably, the knowledge graph can be stored in a Neo4j graph database, so that the question answering system provided by the invention further comprises the Neo4j graph database. The Neo4j graph database is a high-performance NoSQL graph database capable of storing structured data on a network instead of in a table, has the advantages of high performance, strong practicability, light weight and the like, and is also the most commonly used graph database at present. And after preprocessing the collected data, storing the preprocessed data in a triplet form in Neo4j to construct a knowledge graph of rice diseases and insect pests.
Wherein, in the information extraction unit 220, extracting the main entity in the current problem and extracting the target relationship in the current problem through dependency syntax analysis includes: classifying the current problem to obtain the type of the current problem, wherein the type comprises a single relation problem or a complex relation problem; identifying a master entity in the current problem; and extracting the target relation in the current problem according to the type of the current problem through dependency syntax analysis. Wherein, the single relation problem refers to a problem that only a single relation exists, such as: what is the academic name of the rice borer? By the term "only one relationship exists for this problem, which is the academic name of the rice borer. The complex relationship problem is a problem that there are multiple relationships, such as: what is the pest location of the rice grain smut? The problem is a plurality of relations, namely the harmful part of the rice grain smut and the preventing and treating method of the rice grain smut. By classifying the types of the current problems, determining whether the types of the current problems are single relation problems or complex relation problems, after determining, extracting target relations in the current problems according to the types of the current problems in a targeted manner through dependency syntactic analysis, if the types of the current problems are single relation problems, only one relation is considered for extracting the target relations in the current problems, and if the types of the current problems are complex relation problems, the complex relation is considered for extracting the target relations in the current problems, so that the information extraction processing speed is increased, meanwhile, the problem processing is more accurate, and the information extraction efficiency is improved.
Specifically, firstly, a plurality of problem templates related to rice plant diseases and insect pests are constructed and used for classification, wherein the problem templates comprise problem templates with single relations and problem templates with complex relations, so that the range of problem identification can be enlarged. And (3) performing simple entity recognition by using word segmentation to identify each component in the problem. Afterwards, some useless information is deleted, and the sentence is compressed. Meanwhile, dependency syntactic analysis is utilized to identify dependency relations among all components, required relations and entities are extracted, and whether a single relation or a complex relation is extracted is determined according to the problem category. And splicing the finally obtained entity-relation, and carrying out semantic reconstruction to form a new set. Illustratively, for the entity-relation set after semantic reconstruction, a Cypher statement may be written to retrieve the Neo4j graph database to obtain a final answer.
Referring to fig. 3, in the pest and disease relation question-answering page, a user inputs a question about rice pest and disease, then classifies the question, identifies an entity of the question after classifying the question, identifies the entity of the question, compresses a sentence of the question on one hand, and links the entity of the question on the other hand. And carrying out relation extraction after sentence compression, identifying the dependency relation among all components by utilizing dependency syntax analysis, carrying out semantic reconstruction, carrying out retrieval and reasoning on the questions by semantic reconstruction and entity link, and then carrying out retrieval on a Neo4j graph database to generate a final answer.
In summary, in the question-answering system based on the rice pest knowledge graph, the current question input by the user is firstly obtained, and the dependency syntax analysis is carried out on the current question input by the user through the information extraction unit and the retrieval unit, and the reconstruction of semantic information is carried out, so that the technical complexity is reduced. Based on knowledge driving, retrieving and reasoning are carried out on the knowledge graph, so that an answer has interpretability, while an algorithm and a model for deep learning in the prior art have a 'black box effect', so that the transparency of a calculation process is not high, and the method is based on data driving, so that the result lacks interpretability. Therefore, compared with the prior art, the question-answering system based on the rice pest knowledge graph has stronger interpretation. Solves the problem that the answers of the question answering system aiming at the rice diseases and insect pests in the related technology lack the interpretability.
In one embodiment, referring to fig. 4, the question answering system based on the knowledge graph of rice diseases and insect pests further includes: a data processing module; the data processing module includes a data preprocessing unit 410 and a knowledge graph construction unit 420; the data preprocessing unit 410 is used for preprocessing the data of the rice plant diseases and insect pests; the knowledge graph construction unit 420 is configured to perform analysis based on a pre-training language model on the preprocessed rice plant disease and insect pest data, extract entities and relationships therein, and construct the knowledge graph based on the extraction result.
Specifically, the pre-trained language model may employ a MA-RBC model. After preprocessing the collected rice plant disease and insect pest data, the data preprocessing unit performs information extraction on the preprocessed rice plant disease and insect pest data, the extracted entities and relations can be stored in a Neo4j graph database in a triplet form, a rice plant disease and insect pest knowledge graph is built, the triplet is used as a basic unit for describing the entities and relations in the knowledge graph, the relations between the entities can be clearly expressed, and therefore clear semantic association is established, and reply contents can be more accurate. The method comprises the steps of collecting rice pest data, namely collecting rice pest data of different structures from a plurality of channels to the greatest extent, wherein structured data, semi-structured data and unstructured data are covered. Illustratively, structured data may be obtained from a database, semi-structured data may be obtained from web pages of a web site, and unstructured data may be obtained from textual material. For example, the database can be a proprietary database of a professional institute, and the structured data comprises proprietary data applied to the database; the semi-structure data comprises data with certain rules such as agricultural authoritative websites, hundred degrees encyclopedias and the like which are crawled through the Scrapy; unstructured data refers to crawled pure text corpus containing multiple complex relationships and rice pest-related books.
In one embodiment, referring to fig. 5, the question answering system based on the knowledge graph of rice diseases and insect pests further includes: logging in a registration module; the login registration module includes a second interface unit 510, a data specification verification unit 520, and a hash processing unit 530; the second interface unit 510 is configured to obtain login information or registration information input by a user; the data specification checking unit 520 is configured to detect whether the login information or the registration information meets a specification; the hash processing unit 530 is configured to hash-encrypt the registration information when the registration information meets a specification; and the hash verification module is used for carrying out hash verification on the login information when the login information accords with the specification.
Specifically, the second interface unit provides a login and registration page for the user, and the page at least comprises an information input box, so that the user can conveniently input login information and registration information. Illustratively, at the login registration page, components may be rendered using CSS (Cascading Style Sheets), including body, login_box, h1, input_ box, input, button. At the login interface, a user clicks to login, a user name and a password are submitted to a login route by using a POST request of a form, and whether the user name and the password meet requirements or not and whether login is allowed or not is detected by a background. At the login interface, the user clicks to register, uses the GET request of the form, and jumps to the registration page. In the registration page, registration and login return functions are provided, a user clicks registration, a POST request of a form is used for submitting a user name and a password to a register route, and whether the user name and the password meet requirements or not is detected by a background, and whether registration is allowed or not is detected. And detecting whether the user name and the password meet the requirements or not in the login and registration process is realized by a data specification checking unit, and the data specification checking unit is a functional unit in a system background. The hash processing unit performs hash processing on the password by using a function in the werkzeug. In this process, user information is stored by MySQL database, including a user name and a hashed password.
Referring to fig. 6, in the login registration page, if the user does not register, click to register, in the registration page, a registration and return login function is provided, the user input data register, and the background detects whether the data input by the user meets the requirements or not, and whether the registration is allowed or not. If the data input by the user meets the requirements, carrying out hash encryption on the data, registering successfully, returning to a login registration page, otherwise, returning to the registration page after registering failure; if the return login button is clicked, the login registration page is returned again. If the user registers information, the user logs in, the user inputs data, and the background detects whether the data input by the user meets the requirements or not and whether the login is allowed or not. If the data input by the user meets the requirements, carrying out hash verification on the data input by the user, if the verification is successful, entering a first page of a question-answering system based on the rice plant disease and insect pest knowledge graph, otherwise, failing to log in, and returning to a login registration page again; if the data input by the user does not meet the requirements, the login fails and the login registration page is returned again.
In one embodiment, referring to fig. 7, the question answering system based on the knowledge graph of rice diseases and insect pests further includes: a named entity recognition module; the named entity recognition module includes a third interface unit 710, a remote call model unit 720, and a key pair verification unit 730; the third interface unit 710 is configured to obtain a text to be identified input by a user, and output an entity identification result of the text to be identified; the remote calling unit 720 is configured to establish remote connection with a server, send a text to be identified input by a user to the server, call a named entity identification model in the server to identify an entity of the text to be identified, and obtain an entity identification result from the server; the key pair verification unit 730 is used to secure the remote connection through key pair verification.
The third interface unit provides a named entity recognition page to the user, the page having at least a text input box and a text output box. The named entity recognition module can recognize five types of information, namely a main entity, an alias, a place, a harmful part and a pesticide prevention and treatment in the text. Specifically, the remote calling model unit performs entity recognition prediction on the text to be recognized by calling a named entity recognition model on a remote server. For security of the transmission, authentication is performed in the form of a key pair. After the key information is generated locally, public key information is stored in a server side, private key information is stored locally, user verification is carried out through a public key-private key pair, and the verification process is mainly realized by a key pair verification unit. For example, the SSH client may be established using a paramiko library, attached with a remote server IP address, a user name, private key information loaded from a file, and connected to the remote server. A command is then written to activate a particular Conda virtual environment, switch to a model directory, call a model, and send the command to the server through the shell of the SSH. And after the model is activated, transmitting the text to be identified on the page to a server for prediction. The output of the SSH channels is continuously collected during model operation and processed, including elimination of unwanted information, segmentation of data, data formatting, and the like. After the result is obtained, the SSH channel is closed and the connection with the remote server is disconnected. The front end sends text data and receives a prediction result to the back end by using an AJAX request. When the result is received, the front end uses the jQuery framework to display the predicted result at a specific position on the page. The jQuery framework is utilized to simplify DOM operation, and the function of dynamically updating text content when a user interacts with a page is realized.
Preferably, the named entity recognition model adopts an MA-RBC model, and the MA-RBC model comprises a pre-training language layer, a circulating neural network layer, a multi-head self-attention layer and a statistical layer which are sequentially connected. The data to be identified is the input of the MA-RBC model, namely the input of the pre-training language layer, the output of the pre-training language layer is the input of the cyclic neural network layer, the output of the cyclic neural network layer is the input of the multi-head self-attention layer, and the output of the statistical layer forms the output of the MA-RBC model. The MA-RBC model adopted by the invention is a combination of the four layers of networks, and compared with the existing entity identification model, the MA-RBC model is at least added with a multi-head self-attention layer. Multiple head self-attention layers are used in the model to enhance the degree of attention to different locations. In particular, by feature weighting the output of the recurrent neural network layer, different features of attention to multiple attention points can be extracted to better capture important information in the sequence. Firstly, carrying out word coding on marking data through a robust optimized Bode pre-training layer, carrying out text feature extraction on the data to be identified, and dynamically generating word vectors corresponding to single characters according to the contextual features of the words; then inputting the word vector into a two-way long-short-term memory network layer for two-way coding; important information in a text is weighted through a Multi-head self-Attention mechanism in a Multi-head self-Attention layer (Multi-head self-Attention layer), a final text representation is obtained, the final text representation is transmitted into a conditional random field layer for decoding, and a label transfer probability and constraint conditions are obtained through training and learning, so that final prediction labeling data (sequence) is obtained. Experiments are carried out according to the method, and experimental results show that the MA-RBC model adopted by the invention is superior to the existing popular BiLSTM+CRF and BERT+BiLSTM+CRF classical models, and the accuracy is respectively improved by 4.97 percent and 1.8 percent.
Referring to fig. 8, in the named entity recognition page, a user inputs text in a text input box, performs entity recognition prediction on the text to be recognized by calling a named entity recognition model on a remote server, outputs a predicted result in a text output box, displays the predicted result on a web page if the predicted result is available, and does not display the predicted result on the web page if the predicted result is not available.
In one embodiment, referring to fig. 9, the question answering system based on the knowledge graph of rice diseases and insect pests further includes: a plant diseases and insect pests relation retrieval module; the pest and disease relation retrieval module includes a fourth interface unit 910 and an entity relation matching unit 920; the fourth interface unit 910 is configured to obtain an entity to be queried input by a user, and provide an entity selection list for the user, obtain an entity selection instruction of the user for the entity selection list, determine the entity to be queried according to the entity selection instruction, and output a relationship and an entity corresponding to the entity to be queried; the entity relationship matching unit 920 is configured to match a relationship corresponding to an entity to be queried with an entity in the knowledge graph.
Specifically, the fourth interface unit provides a pest and disease relationship retrieval page for the user. Preferably, the pest and disease relation retrieval page is imported with files such as bootstrap. Min. Css, icons. Js, echartits. Min. Js and the like, so that page rendering and page interaction are facilitated. The pest and disease relation retrieval page at least comprises a text input box, a label list and an image display area. Both the text entry box and the tab list may view a relationship graph of pest entities. The tag list contains limited plant diseases and insect pests names, has a dynamic display effect, and can be used for clicking and checking the entity by a user through a mouse to achieve the same effect as that of searching by inputting texts in a text input box. For example, the ECharts framework may be used to draw a relationship graph at the web page end to represent the corresponding entity-relationship. The front end transmits the name of the plant diseases and insect pests to be inquired to the rear end as parameters, and the rear end obtains the parameters through the requests. The relation and the entity related to the entity can be queried by using a map query language Cypher based on Neo4j, the query matching process is realized by an entity relation matching unit, and the entity relation matching unit belongs to a back-end functional unit. In order to fully reveal the relation diagram, the queried data needs to be converted into a JSON (JavaScript Object Notation) format, which is a lightweight data exchange format, is easy to read and write by people and is easy to analyze and generate by machines. The relation diagram mainly comprises nodes and edges, so that a key value pair is needed in the JSON file to store node data, namely the inquired entity information; a key pair is required to store edge data, i.e., the relationship between entities. After the front end receives the JSON data processed by the back end, the relationship diagram in ECharts is correspondingly set, and the relationship diagram corresponding to the entity is drawn. And a zoom function is added in the display area, so that a user can more intuitively feel data by zooming in or out the relation diagram through a mouse wheel.
Referring to fig. 10, in the pest and disease relation search page, a pest and disease name to be searched is input in a text input box, a search button is clicked, the front end transmits the pest and disease name to be searched as a parameter to the rear end, the rear end obtains the parameter through requests, searches in a Neo4j graph database, and matches the parameter with the entity; if the button in the tag list is directly clicked, the same effect is that the front end transmits the name of the plant diseases and insect pests to be queried to the rear end as a parameter, and the rear end acquires the parameter through the requests, searches in the Neo4j graph database and matches the parameter with the entity.
In one embodiment, referring to fig. 11, the question answering system based on the knowledge graph of rice diseases and insect pests further includes: a rice plant disease and insect pest visualization module; the rice plant disease and insect pest visualization module comprises a fifth interface unit 1110 and a visualization unit 1120; the fifth interface unit 1110 is configured to obtain a zoom instruction input by a user; the visualization unit 1120 is configured to display the knowledge graph to a user, and scale the knowledge graph according to the scaling instruction.
Specifically, the fifth interface unit provides a visual page for rice diseases and insect pests for the user, and the page at least comprises a knowledge graph display area and a global display button; the visualization unit converts the data format of the knowledge spectrogram in the database into a data format convenient for presentation, such as a JSON file. Thereby displaying the knowledge spectrum of the rice plant diseases and insect pests in the knowledge spectrum display area. Illustratively, the knowledge-graph presentation area displays 150 relationships and related entities by default. Clicking the global display button can display the complete knowledge graph in the Neo4j graph database, and if the global display button is not clicked, displaying part of data. The knowledge graph is displayed by a relation graph in the ECharts framework. Before drawing the relationship graph, a JSON file of knowledge-graph related data needs to be constructed. And respectively inquiring 150 records and all records in Neo4j through a Cypher language, and converting the records into a JSON file with node data and edge data. The front end sends out a request, obtains JSON data, then sets a relation diagram including a title, a font, a layout and the like, imports the JSON data, and updates the diagram to display a corresponding knowledge graph. The knowledge-graph presentation area also has a scalability function.
Referring to fig. 12, when entering a visual page of rice plant diseases and insect pests for the first time, the front end sends a request to acquire JSON data, the system defaults to read part of JSON data, and then ECharts are configured to draw a knowledge graph; clicking a global display button, sending a request from the front end, acquiring JSON data, reading all the JSON data by a system, and then configuring ECharts to draw a knowledge graph.
As above, explanation analysis has been made on the question-answering system based on the knowledge graph of rice plant diseases and insect pests by various embodiments. As can be seen from the above, referring to fig. 13, the question answering system based on the rice pest knowledge graph provided by the invention performs hash processing on the password in the login and registration process in the login and registration module, so as to increase the security of data storage. In the named entity recognition module, named entity recognition is carried out on the text by remotely calling the MA-RBC model, so that the recognition rate is high. And in the plant disease and insect pest relation retrieval module, the retrieval of the related plant disease and insect pest relation is realized based on the constructed knowledge graph. And in the rice plant diseases and insect pests visualization module, an ECharts framework is adopted to realize the visualization function of the data in the Neo4j graph database. In the disease and pest relation question-answering module, the dependency syntax analysis is utilized to analyze the question structure, a single relation or a complex relation is extracted, and the retrieval and reasoning are carried out by means of an answer generation algorithm. In the practical application process, the data is preprocessed, and then a Neo4j graph database is utilized to construct a rice pest knowledge graph for relevant data conversion. Thereby solving the problem of lack of interpretability in general question-answering systems (e.g. based on data driving)
In some embodiments, a question-answer test is performed on a question-answer system based on a rice pest knowledge graph, as shown in table 1, the test questions and corresponding predicted results and actual answers are shown, and it can be seen from the table that the actual answers are consistent with expectations.
TABLE 1 Pest relationship question-answer test
The above-described respective modules and units may be functional modules and units, or program modules and units, and may be implemented by software or hardware. For modules implemented in hardware, the various modules described above may be located in the same processor; or the above modules may be located in different processors in any combination.
The invention also provides a question answering method based on the rice pest knowledge graph, which comprises the following steps: acquiring a current problem input by a user; extracting a main entity in the current problem, and extracting a target relation in the current problem through dependency syntax analysis; and searching a target guest entity with a target relation with the host entity in the pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
In the question-answering method based on the rice pest knowledge graph, the current question input by the user is acquired, the dependency syntax analysis is carried out on the current question input by the user, the semantic information is reconstructed, the technical complexity is reduced, and the knowledge graph is searched and inferred based on knowledge driving, so that the answer has interpretability. Whereas the algorithms and models of deep learning in the prior art have a "black box effect" which results in a low transparency of the calculation process, such methods are data driven and lack interpretability of the results. Therefore, compared with the prior art, the question-answering method based on the rice pest knowledge graph has stronger interpretation. Solves the problem that the answers of the question answering system aiming at the rice diseases and insect pests in the related technology lack the interpretability.
In some of these embodiments, for the information extraction unit 220, extracting the master entity in the current question, and for extracting the target relationship in the current question through dependency syntax analysis, includes:
classifying the current problem to obtain the type of the current problem, wherein the type comprises a single relation problem or a complex relation problem; identifying a master entity in the current problem; and extracting the target relation in the current problem according to the type of the current problem through dependency syntax analysis.
By classifying the types of the current problems, determining whether the types of the current problems are single relation problems or complex relation problems, after determining, extracting target relations in the current problems according to the types of the current problems in a targeted manner through dependency syntactic analysis, if the types of the current problems are single relation problems, only one relation is considered for extracting the target relations in the current problems, and if the types of the current problems are complex relation problems, the complex relation is considered for extracting the target relations in the current problems, so that the information extraction processing speed is increased, meanwhile, the problem processing is more accurate, and the information extraction efficiency is improved.
It should be noted that the steps illustrated in the above-described flow or flow diagrams of the figures may be performed in a computer system, such as a set of computer-executable instructions, and that, although a logical order is illustrated in the flow diagrams, in some cases, the steps illustrated or described may be performed in an order other than that illustrated herein.
There is also provided in the invention an electronic device comprising a memory in which a computer program is stored and a processor arranged to run the computer program to perform the steps of any of the method embodiments described above.
Optionally, the electronic apparatus may further include a transmission device and an input/output device, where the transmission device is connected to the processor, and the input/output device is connected to the processor.
Alternatively, in one embodiment, the processor may be arranged to perform the following steps by a computer program:
s1, acquiring a current problem input by a user;
s2, extracting a main entity in the current problem, and extracting a target relation in the current problem through dependency syntax analysis;
and S3, searching a target guest entity with a target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
It should be noted that, the specific examples of the present electronic device may refer to examples described in the embodiments and the optional implementations of the method, and are not described in detail in this embodiment.
In addition, in combination with the question-answering method based on the rice pest knowledge graph, which is provided by the invention, a storage medium can be provided for realizing the question-answering method. The storage medium has a computer program stored thereon; the computer program when executed by the processor realizes any question answering method based on the rice pest knowledge graph in the embodiment.
It should be noted that, user information (including but not limited to user equipment information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, presented data, etc.) referred to in the present application are information and data authorized by the user or sufficiently authorized by each party.
It should be understood that the specific embodiments described herein are merely illustrative of this application and are not intended to be limiting. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present application, are within the scope of the present application in light of the embodiments provided herein.
It is evident that the drawings are only examples or embodiments of the present application, from which the present application can also be adapted to other similar situations by a person skilled in the art without the inventive effort. In addition, it should be appreciated that while the development effort might be complex and lengthy, it would nevertheless be a routine undertaking of design, fabrication, or manufacture for those of ordinary skill having the benefit of this disclosure, and thus should not be construed as an admission of insufficient detail.
The term "embodiment" in this application means that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the application. The appearances of such phrases in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive. It will be clear or implicitly understood by those of ordinary skill in the art that the embodiments described in this application can be combined with other embodiments without conflict.
Claims (10)
1. A question-answering system based on a knowledge graph of rice diseases and insect pests, the system comprising: a disease and pest relation question-answering module;
the pest and disease relation question-answering module comprises a first interface unit, an information extraction unit and a retrieval unit;
the first interface unit is used for acquiring a current problem input by a user and outputting a reply corresponding to the current problem;
the information extraction unit is used for extracting a main entity in the current problem and extracting a target relation in the current problem through dependency syntax analysis;
the searching unit is used for searching a target guest entity with the target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
2. The question-answering system based on the knowledge spectrum of rice plant diseases and insect pests according to claim 1, wherein the knowledge spectrum is stored in a Neo4j graph database.
3. The rice pest knowledge graph-based question-answering system according to claim 1, further comprising: a data processing module;
the data processing module comprises a data preprocessing unit and a knowledge graph construction unit;
the data preprocessing unit is used for preprocessing the rice plant diseases and insect pests data;
the knowledge graph construction unit is used for analyzing the preprocessed rice plant diseases and insect pests data based on a pre-training language model, extracting entities and relations in the preprocessed rice plant diseases and insect pests data, and constructing the knowledge graph based on extraction results.
4. The rice pest knowledge graph-based question-answering system according to claim 1, further comprising: logging in a registration module;
the login registration module comprises a second interface unit, a data specification verification unit and a hash processing unit;
the second interface unit is used for acquiring login information or registration information input by a user;
the data specification checking unit is used for detecting whether the login information or the registration information accords with a specification;
The hash processing unit is used for carrying out hash encryption on the registration information when the registration information accords with the specification; and the hash verification is used for carrying out hash verification on the login information when the login information accords with the specification.
5. The rice pest knowledge graph-based question-answering system according to claim 1, further comprising: a named entity recognition module;
the named entity identification module comprises a third interface unit, a remote call model unit and a key pair verification unit;
the third interface unit is used for acquiring a text to be identified input by a user and outputting an entity identification result of the text to be identified;
the remote calling unit is used for establishing remote connection with a server, sending a text to be identified, which is input by a user, to the server, calling a named entity identification model in the server to identify the entity of the text to be identified, and acquiring an entity identification result from the server;
the key pair verification unit is used for guaranteeing the security of remote connection.
6. The rice pest knowledge graph-based question-answering system according to claim 1, further comprising: a plant diseases and insect pests relation retrieval module;
The plant diseases and insect pests relation retrieval module comprises a fourth interface unit and an entity relation matching unit;
the fourth interface unit is used for acquiring an entity to be queried input by a user, providing an entity selection list for the user, acquiring an entity selection instruction of the user for the entity selection list, determining the entity to be queried according to the entity selection instruction, and outputting a relation and an entity corresponding to the entity to be queried;
the entity relation matching unit is used for matching the relation and the entity corresponding to the entity to be queried in the knowledge graph.
7. The rice pest knowledge graph-based question-answering system according to claim 1, further comprising: a rice plant disease and insect pest visualization module;
the rice plant diseases and insect pests visualization module comprises a fifth interface unit and a visualization unit;
the fifth interface unit is used for acquiring a zoom instruction input by a user;
the visualization unit is used for displaying the knowledge graph to a user and scaling the knowledge graph according to the scaling instruction.
8. The rice pest knowledge graph-based question-answering system according to claim 5, wherein the named entity recognition model comprises a MA-RBC model.
9. The rice pest knowledge graph-based question-answering system according to claim 1, wherein the extracting the main entity in the current question extracts the target relationship in the current question through dependency syntax analysis, comprising:
classifying the current problem to obtain the type of the current problem, wherein the type comprises a single relation problem or a complex relation problem;
identifying a master entity in the current question;
and extracting the target relation in the current problem according to the type of the current problem through dependency syntax analysis.
10. A question answering method based on a rice pest knowledge graph is characterized by comprising the following steps:
acquiring a current problem input by a user;
extracting a main entity in the current problem, and extracting a target relation in the current problem through dependency syntax analysis;
and searching a target guest entity with the target relation with the host entity in a pre-constructed knowledge graph, and generating a reply of the current problem according to the target guest entity.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311763394.1A CN117436531A (en) | 2023-12-21 | 2023-12-21 | Question answering system and method based on rice pest knowledge graph |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202311763394.1A CN117436531A (en) | 2023-12-21 | 2023-12-21 | Question answering system and method based on rice pest knowledge graph |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN117436531A true CN117436531A (en) | 2024-01-23 |
Family
ID=89546482
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202311763394.1A Pending CN117436531A (en) | 2023-12-21 | 2023-12-21 | Question answering system and method based on rice pest knowledge graph |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN117436531A (en) |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112231460A (en) * | 2020-10-27 | 2021-01-15 | 中国科学院合肥物质科学研究院 | Construction method of question-answering system based on agricultural encyclopedia knowledge graph |
| US20220391426A1 (en) * | 2022-03-02 | 2022-12-08 | Beijing Baidu Netcom Science Technology Co., Ltd. | Multi-system-based intelligent question answering method and apparatus, and device |
| CN116010564A (en) * | 2022-12-14 | 2023-04-25 | 扬州大学 | Construction method of rice pest question-answering system based on multi-mode knowledge graph |
| CN116257610A (en) * | 2023-01-11 | 2023-06-13 | 山西长河科技股份有限公司 | Intelligent question-answering method, device, equipment and medium based on industry knowledge graph |
| CN116306652A (en) * | 2023-03-15 | 2023-06-23 | 广东工业大学 | Chinese naming entity recognition model based on attention mechanism and BiLSTM |
| CN116450785A (en) * | 2023-03-06 | 2023-07-18 | 中国电子科技集团公司第二十八研究所 | Knowledge question-answering method based on flight situation knowledge graph |
| CN116719915A (en) * | 2023-05-15 | 2023-09-08 | 中国工商银行股份有限公司 | Intelligent question-answering method, device, equipment and storage medium |
| CN117033571A (en) * | 2023-06-27 | 2023-11-10 | 山东新一代信息产业技术研究院有限公司 | Knowledge question-answering system construction method and system |
| CN117076688A (en) * | 2023-08-18 | 2023-11-17 | 中国工商银行股份有限公司 | Knowledge question-answering method and device based on domain knowledge graph and electronic equipment |
| CN117198558A (en) * | 2022-05-27 | 2023-12-08 | 上海说以科技有限公司 | System and method for evaluating pharmaceutical project based on knowledge graph |
-
2023
- 2023-12-21 CN CN202311763394.1A patent/CN117436531A/en active Pending
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112231460A (en) * | 2020-10-27 | 2021-01-15 | 中国科学院合肥物质科学研究院 | Construction method of question-answering system based on agricultural encyclopedia knowledge graph |
| US20220391426A1 (en) * | 2022-03-02 | 2022-12-08 | Beijing Baidu Netcom Science Technology Co., Ltd. | Multi-system-based intelligent question answering method and apparatus, and device |
| CN117198558A (en) * | 2022-05-27 | 2023-12-08 | 上海说以科技有限公司 | System and method for evaluating pharmaceutical project based on knowledge graph |
| CN116010564A (en) * | 2022-12-14 | 2023-04-25 | 扬州大学 | Construction method of rice pest question-answering system based on multi-mode knowledge graph |
| CN116257610A (en) * | 2023-01-11 | 2023-06-13 | 山西长河科技股份有限公司 | Intelligent question-answering method, device, equipment and medium based on industry knowledge graph |
| CN116450785A (en) * | 2023-03-06 | 2023-07-18 | 中国电子科技集团公司第二十八研究所 | Knowledge question-answering method based on flight situation knowledge graph |
| CN116306652A (en) * | 2023-03-15 | 2023-06-23 | 广东工业大学 | Chinese naming entity recognition model based on attention mechanism and BiLSTM |
| CN116719915A (en) * | 2023-05-15 | 2023-09-08 | 中国工商银行股份有限公司 | Intelligent question-answering method, device, equipment and storage medium |
| CN117033571A (en) * | 2023-06-27 | 2023-11-10 | 山东新一代信息产业技术研究院有限公司 | Knowledge question-answering system construction method and system |
| CN117076688A (en) * | 2023-08-18 | 2023-11-17 | 中国工商银行股份有限公司 | Knowledge question-answering method and device based on domain knowledge graph and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11501182B2 (en) | Method and apparatus for generating model | |
| WO2021082953A1 (en) | Machine reading understanding method and apparatus, storage medium, and device | |
| US20200301954A1 (en) | Reply information obtaining method and apparatus | |
| CN107679039B (en) | Method and device for determining statement intention | |
| CN111125309A (en) | Natural language processing method and device, computing equipment and storage medium | |
| CN111090987A (en) | Method and apparatus for outputting information | |
| CN116775847B (en) | Question answering method and system based on knowledge graph and large language model | |
| CN112650840A (en) | Intelligent medical question-answering processing method and system based on knowledge graph reasoning | |
| CN112214593A (en) | Question and answer processing method and device, electronic equipment and storage medium | |
| CN112287069B (en) | Information retrieval method and device based on voice semantics and computer equipment | |
| CN113704460B (en) | Text classification method and device, electronic equipment and storage medium | |
| CN111783903B (en) | Text processing method, text model processing method and device and computer equipment | |
| CN115062134B (en) | Knowledge question-answering model training and knowledge question-answering method, device and computer equipment | |
| US20240152770A1 (en) | Neural network search method and related device | |
| CN113761220A (en) | Information acquisition method, device, equipment and storage medium | |
| CN114360731A (en) | Method for constructing medical care and knowledge combined service system for old people care | |
| CN113704420A (en) | Method and device for identifying role in text, electronic equipment and storage medium | |
| CN117093687A (en) | Question answering method and device, electronic equipment and storage medium | |
| CN113342944B (en) | Corpus generalization method, apparatus, device and storage medium | |
| CN112199954B (en) | Disease entity matching method and device based on voice semantics and computer equipment | |
| CN111403011B (en) | Registration department pushing method, device and system, electronic equipment and storage medium | |
| CN114372454A (en) | Text information extraction method, model training method, device and storage medium | |
| CN114491076B (en) | Data enhancement method, device, equipment and medium based on domain knowledge graph | |
| CN117436531A (en) | Question answering system and method based on rice pest knowledge graph | |
| CN115130461A (en) | Text matching method and device, electronic equipment and storage medium |
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 |