Design And Implementation Of Intelligent Question Answering System Based On Knowledge Map Of Agricultural Diseases And Insect Pests

Inquiry and knowledge acquisition of agricultural diseases and insect pests is a crucial part of the planting process for farmers.However,it is difficult for farmers to consult experts and communicate less with peers in the consultation of diseases and insect pests.The appearance of question and answer system brings good news to farmers.Therefore,this paper developed an intelligent question answering system based on the knowledge map of agricultural diseases and insect pests,which provides users with more accurate,convenient and comprehensive answering service of agricultural diseases and insect pests knowledge through the knowledge map of agricultural diseases and insect pests.This paper focuses on the design and implementation of intelligent question answering system for agricultural diseases and insect pests and carried out the following work:(1)Analyze the requirements,feasibility,hierarchical structure and constituent structure of the intelligent question answering system of agricultural diseases and insect pests.Firstly,the basic functions of the system and factors affecting user experience are analyzed from functional requirements and nonfunctional requirements.Then,the feasibility of the system is judged from two aspects of technology and economy.Finally,from the analysis of the design principle and function of system level architecture and the composition of the system structure,among them,the level of the system architecture including frontend interactive layer and server layer,layer,front and back end structure includes four modules,respectively,knowledge map module,human-computer interaction module,understand and answer matching module.(2)The knowledge map of agricultural diseases and insect pests was constructed.Firstly,a large number of structured data of pests and diseases were obtained from the related websites of pests and diseases by web crawler technology,and sorted into the form of "entity-relationship-attribute" triplet through data cleaning.Then,using Neo4 j graph database as storage medium,triples were imported into graph database to construct agricultural disease and insect knowledge map.The knowledge map contains660 agricultural pest and disease entities,1937 relationships and about 4000 related attribute values.It can clearly represent the relationship between entities and attributes,and has great advantages in data storage and data query.(3)Aiming at the problem of insufficient understanding and inaccurate answers in traditional QUESTION answering system,a method combining Chinese word segmentation,part of speech substitution,naive Bayes and knowledge graph is proposed to realize the function of question understanding and answering in intelligent question answering system.Firstly,the problem of natural language form is processed by word segmentation,and the problem template is matched by naive Bayes classification algorithm.Then,the abstract result of the problem is obtained by part-of-speech substitution.Finally,the abstract results are combined with Cypher language to query the corresponding answers.The main functions of the system are realized through the combination of several technologies,and the accuracy of the correct answer reaches 93%.(4)In view of the problems of untidy code and inconvenient management of traditional back-end development technology,front-end applications need to be downloaded and inconvenient to use,a system development method of Spring-boot combined with wechat applets is proposed.Firstly,the back end of intelligent question answering system is developed by Spring-boot.Then,using wechat small program to complete the front-end development of the system,to achieve the clear screen,problem prompts and other functions,and through WXML and WXSS language to achieve the front-end page beautification.The combination of the above two technologies improves the development efficiency and reduces the user’s difficulty in using them.(5)The function test and pressure test of the system were carried out.Firstly,the test question sets are input into the system and the output is compared with the correct answers.The results show that the correct answer rate of the system reaches 93%.The results show that the average response time is less than 1 second,the maximum utilization rate of hardware CPU is 75%,the maximum utilization rate of memory is 68%,and all the performance indicators are good,which can meet the requirements of normal operation.The establishment of the intelligent question and answer system of agricultural diseases and insect pests provides a platform for farmers to inquire the knowledge of agricultural diseases and insect pests at any time,enriches the means of disease and insect pests prevention and control of farmers,and promotes the popularization of the knowledge of agricultural diseases and insect pests.