Research And Application Of High School Mathematical Knowledge Atlas Based On Ontology Learning

With the rapid development of educational informationization in the 21 st century,computer-assisted instruction(CAI)is a kind of teaching form highly recommended by the country and society.However,due to the miscellaneous network learning resources,it is easy for middle school students to get lost in knowledge and overload knowledge when learning without the guidance of teachers.Therefore,the construction of knowledge map of high school mathematics is a basic work of computer-aided teaching and can become an important tool for teachers to conduct teaching guidance in the teaching process.To establish a high school mathematics knowledge map,it is necessary to carry out ontology learning on the relationship between concepts and concepts,so as to extract knowledge that can accurately describe the definitions and axioms of high school mathematics.This paper mainly extracts concepts and concepts from the definitions and theorems of high school mathematics knowledge from Baidu Encyclopedia and Chinese Wikipedia,thus forming a high school math ontology knowledge base.The main work is as follows:(1)Linguistic-based concept extraction Although the corpus is analyzed according to semantic grammar,there is little ambiguity,but the efficiency of processing large amounts of corpus data is low;The concept extraction based on statistics extracts concepts according to statistical information such as word frequency.Since there is no semantic grammar support,the accuracy of extraction is not high.In this paper,the two methods are combined.Firstly,the part-of-speech rule base is established.According to the rule of word combination,matching in the rule base is used to obtain the concept candidate set in the high school mathematics knowledge field.Then,based on the concept candidate set,the statistical algorithm combined with mutual information,left and right information entropy and TF-IDF algorithm is used to screen it,and the concept set of high school mathematics knowledge field is obtained.(2)Aiming at the parent-child relationship and other classification relations,firstly,the page classification structure of the encyclopedia entry was clustered,and then the concept parent-child pairs were obtained through co-occurrence analysis.It was found that when the error rate was 5%,the number of concept parent-child pairs was large and the threshold value was reasonable.For the non-classification relationship,the concept pairs in the candidate concept set are extracted by the association rule.After the experiment,it is found that when the support threshold is 0.000197 and the confidence threshold is 0.0103,the extraction effect is better,and the concept pair is obtained 319 pairs;Then use the SCWS word segmentation system to obtain the predicate verbs,and use the TF-IDF algorithm to filter the company to obtain the predicate verbs that can represent the relationship.Finally,the triad model can be constructed according to the previously extracted concept father and son pairs and predicate verbs.The concept parent-child pair is matched with the predicate verb according to the likelihood function.This paper constructs the ontology knowledge base of high school mathematics by extracting the concept of concept and concept between Chinese high school mathematics and high school mathematics,and uses the Protégé tool to visualize the high school mathematics knowledge into a knowledge map and apply it to the system to make it more Practical.