Research On The Construction And Application Of Knowledge Graph For Middle School Mathematics Subject Based On Neural Network

With the continuous progress of information technology,various applications in the field of education begin to tend to be information-based and intelligent.In the context of the rapid development of artificial intelligence and big data,the advantages of knowledge graphs in organizing and displaying disciplinary architecture,analyzing and mining disciplinary knowledge associations have become increasingly prominent.Especially in the field of wisdom education,the construction of subject knowledge map and the application of innovation have become the current research hotspot,which has very important research significance and application value.In view of the needs in the field of mathematics,this thesis makes a comprehensive study on the construction and application of mathematics knowledge map in junior middle school.The specific research work includes the following aspects:(1)Data acquisition and preprocessing of junior middle school mathematics subject knowledge.The original data of junior high school mathematics subject knowledge obtained in this paper mainly come from the text content of junior high school mathematics textbook and network resources obtained based on Scrapy crawler technology.Then,data cleaning,word segmentation and part-of-speech tagging are preprocessed successively,subsequent data set construction for junior middle school mathematics subject knowledge extraction is completed.(2)Research on unstructured text-oriented junior middle school mathematics knowledge extraction method.For the triplet extraction task in unstructured text,firstly,this thesis proposes a NER model based on Bi LSTM-CRF.Secondly,a relation extraction model based on Attention Guided Graph Convolutional Network(AGGCN)and a text classification model based on GCN are proposed by introducing Graph Convolutional Network(GCN)and attention mechanism respectively which realize the automatic extraction of junior high school mathematical knowledge triples.The experimental results show that these models can effectively complete the extraction of junior high school mathematics knowledge.(3)The knowledge graph consists of a schema layer and a data layer.In view of the multi-source data obtained from different platforms,combined with the characteristics of junior high school mathematics.First,this thesis designs an ontology construction process suitable for junior high school mathematics and completes the schema layer construction of junior high school mathematics subject knowledge map.Secondly,the data layer of the knowledge map of the junior high school mathematics subject is constructed and realized by the junior high school mathematics knowledge extraction model.Finally,this thesis uses the Neo4 j graph database to store and visualize triples.(4)The knowledge query platform for junior high school mathematics subject knowledge map is constructed.Based on the constructed knowledge graph,this thesis er develops a visual query platform for junior high school mathematics knowledge.The platform not only integrates core functions such as named entity recognition,relation extraction,and text classification,but also provides knowledge visualization queries and retrieval based on a knowledge graph.It has certain practical value.