Research And Implementation Of Knowledge Map And Application Aid App For Apple Pests And Diseases

Apple is a genus of apples in the subfamily Appleaceae of the Rosaceae family,and China is the country with the largest apple cultivation area and highest production.The apple fruit is crisp,sweet and rich in vitamins and minerals,and rich in calcium,making it one of the world’s top four fruits.In the apple growing process,apple pests and diseases are the key factors that affect the appearance,taste and nutritional value of apples,thus affecting the yield,quality,grade and food value of apples.However,most of the existing apple pest knowledge is stored in traditional books and scattered Internet data,which has problems such as long query time,low knowledge utilization,redundant search results,and poor query experience.To address the above problems,this study takes apple pests and diseases as the research object,establishes apple pest and disease data sets,constructs apple pest and disease knowledge graphs,and designs and implements an apple pest and disease application aid app,and the specific research work and conclusions are as follows:(1)Knowledge graph construction for apple pests and diseases.In view of the characteristics of existing apple pests and diseases,a top-down approach is adopted to construct the knowledge map,and the construction process is divided into four steps: data set establishment,schema layer design,knowledge extraction and knowledge map construction and visualization.Firstly,according to the different characteristics of the data,different methods are adopted to obtain the data and establish the apple pest and disease data set.Then pre-processing operations such as cleaning,filtering and rejecting the duplicate data and invalid data present in it are performed,and BIOES sequence tagging strategy is used to tag the information so that it can be understood by computers.Then the knowledge in the field of apple pests and diseases is deeply analyzed,a reasonable knowledge graph pattern layer is designed,and the information that meets the requirements is extracted from the existing data to obtain the apple pest and disease triad.Finally,the Py2 neo library in Python is used to associate with the Neo4 j graph database to complete the automated creation and visualization of the apple pest knowledge graph,forming a knowledge graph with809 nodes,1224 relations,12 types of labels and 10 types of relations.In addition,the issues of data correctness,accuracy and validity were considered in the process of building the knowledge graph to improve the quality and credibility of the apple pest and disease knowledge graph.(2)Joint extraction of apple pest and disease entities based on BERT-Bi LSTM-CRF.For the problem of fragmented information in apple pest and disease domain,this paper adopts the joint extraction method of apple pest and disease entities based on BERT-Bi LSTM-CRF for experiments.First,the language model is trained on the original apple pest and disease dataset,and the character-based semantic vector is generated using the bi-directional Transformer network structure of the BERT part.Then the word vectors are used as the input of Bi LSTM to obtain the association information between entities.Finally,the CRF model is used to obtain the optimal token sequence,thus improving the accuracy and recall of entity joint extraction.After training,the accuracy of the model is 94.36%,the recall rate is 71.69%,and the F1 value is 81.40%,indicating that the method has good performance in the field of joint entity extraction of apple pests and diseases,and has some practical application value for natural language processing and knowledge graph construction.(3)Design and implementation of apple pest and disease application aid App.Based on the knowledge map of apple pests and diseases,we analyzed the requirements according to the actual situation and completed the system function design,combining HTML5,CSS3,Java Script,Vue,Spring Boot and other technologies to realize several functional modules such as disease encyclopedia,application measures,intelligent Q&A and personal center.The question and answer function receives natural language questions from users about apple pests and diseases,analyzes the intent of the questions and finds and returns semantic answers in the knowledge graph,so as to quickly answer users’ questions.The app can provide users with information and consulting services to assist them in making decisions on apple pest and disease control to achieve sustainable high efficiency,high quality and high yield of apples.