Key Technology And Application Of Understanding Elementary Mathematical Problem Based On Word Embedding

In recent years,with the development of artificial intelligence,all aspects of life have become more intelligent.In the field of education,how to make the power of artificial intelligence better serve the human education is a very meaningful subject.To allow the computer to personalize counseling and automatic problem solving,the first thing to solve is to let the computer understand the mathematical language used by humans in order to achieve computer automatic reasoning and automatic problem solving.This paper mainly studies the key methods of comprehension in the field of elementary mathematics,including the characteristics of elementary mathematics language and its knowledge representation method,the representation of words in elementary mathematics,the feature extraction of mathematical short texts and the method of feature construction,Comprehension process and the preprocessing method of mathematical text.Because elementary mathematics language is different from the particularity of ordinary natural language,it is impossible to directly use the existing method of dealing with ordinary natural language by studying the mathematical language of elementary mathematics.Compared with the ordinary natural language,the mathematical language contains a large number of symbols,formulas,proper nouns and other ordinary natural language does not have the content,so to do the primary mathematical sense of the problem must be based on these characteristics of the ordinary natural language processing methods Make improvements.The main work of this paper is as follows:1?the characteristics of elementary mathematics language are analyzed in detail,and a simple and efficient method of elementary mathematics knowledge representation is designed according to its characteristics and design.1)In the representation of the word,this paper deeply studies the model and algorithm of the open source tool word2 vec which has achieved good effect in the natural language field.Then,on the basis of the characteristics of the mathematical language,it puts forward the key to the field of mathematics Word weighting,and so on,making it more suitable for expressing words in mathematical languages.2)At the presentation level of the sentence,the short text is represented by the combination of the semantic dimension and the two features of the mathematical entity.The synergistic expression of the two features makes it more sufficient to express the characteristics of the mathematical short sentences and is more conducive to the follow-up work.2?In the way to understand the meaning of the text,this article another way to understand the problem of mathematical problems into the classification of mathematical essay problem,making the problem simple,according to the existing method can be a good solution to the unknown problem.In the classification of short text of mathematics,according to the characteristics of semantic and mathematical cooperative feature vector,this paper combines two kinds of existing classification algorithms and improves them,and proposes a classification algorithm suitable for classification of mathematical text.In the specific process of comprehension comprehension,we first design the text preprocessing method for the primary mathematics field,then construct the eigenvector through the word vector and the mathematical entity,then classify the short text of the mathematics,and finally generate the corresponding knowledge representation.And the realization of the system used in the initial mathematical problems automatically solve the system.Based on the above methods,this paper designs and implements an elementary mathematical problem understanding system.And through comparative experiments to prove that the system in the primary mathematical understanding of the problem has achieved good results.