Research And Realization On Automatic Solving Of Sequence Problem In Humanoid Answering System

In the era of rapid development of artificial intelligence,more and more researchers began to pay attention to the combination of artificial intelligence and education.Mathematics is the foundation of all science and engineering,and mathematics is relatively "rational" in terms of expression,which facilitates machine understanding.Therefore,artificial intelligence applications will be a good choice for mathematics education.The research content of this topic is the automatic reasoning based on artificial intelligence,on the basis of the correct understanding(NLP),to solve the problem of elementary sequence automatically,and to output the humanoid answer steps.The research results of this paper can be applied to mathematics education,such as counseling students to answer questions,marking test points for the subjects,and helping teachers to correct their assignments.Therefore,this paper is of great significance for mathematics education.The content of this paper is as follows:The research and design of the knowledge representation of the sequence.What knowledge representation do is to help the computer build the related concepts in a certain field.Just as human beings want to solve a mathematical problem in a certain way,we need to know some of the mathematical concepts in this area.In the knowledge representation of sequence,we define the concept of Java class to describe the related concepts of sequence.The research of automatic reasoning for sequence.Automatic reasoning of sequence includes rule reasoning and auxiliary reasoning.For rule reasoning,I have used the rules and formulas of the series of mathematics in the elementary mathematics textbooks as rules,and placed them in the rule-based reasoning library.At the same time,according to the characteristics of the rules,the rules are subdivided into calculation rules,logic rules and intuition rules.For the auxiliary reasoning,it needs to be completed by the symbol calculation engine maple.This is the core part of the series problem solving.The implementation of auto-halt.After an answer is solved by a computer,the reasoning engine needs to be aware of the need to stop reasoning.This part of the work is realized by the auto-halt module.The realization of auto-halt depends on the establishment of a halt rule base.At the same time,auto-halt may occur at the wrong time,resulting in incorrect answers and even unanswered questions,which will bementioned in this paper.The research and realization of humanoid answers.After answering each question,the system will automatically generate humanoid answers.The generation algorithm of the humanoid answer process is essentially a post-order traversal algorithm of a tree.How to determine which knowledge in the knowledge base must appear in the steps of the answer,how to find such knowledge,these issues are the content of this paper.After completing the above research content,the original humanoid solving system has a good solution to the problem of elementary mathematics sequence.At present,the solution rate of the system to the sequence questions reaches 71%,and the average time to solve one problem is less than 3 seconds.