Knowledge Discovery Based On Inductive Learning And Its Application In Automatic Problem Solving System

In recent years,many experts and scholars at home and abroad have devoted a lot of energy to the research of artificial intelligence,which has promoted the rapid development of artificial intelligence in various fields.Among them,artificial intelligence in the field of education will be an important direction of artificial intelligence research,and mathematics is the most difficult fortress to overcome in the field of education.Previously,our team has studied and developed an AutomaticProblem-Solving-System for elementary mathematics,but the accuracy and efficiency of problem-solving can not meet the needs of education.The aim of this paper is to improve the accuracy and efficiency of problem solving by combining inductive learning with automatic problem solving system and assisting the reasoning of automatic problem solving system.The main research contents are as follows:Analysis and mastery of the structure of Automatic-Problem-Solving-System.The Automatic-Problem-Solving-System is the object of knowledge discovery application based on inductive learning.Therefore,it is necessary to analyze the structure of the Automatic-Problem-Solving-System and master the reasoning mode and deep-seated principle of the automatic problem-solving system before considering how to apply inductive learning to assist reasoning.This paper will analyze the Automatic-Problem-Solving-System from three aspects: system structure,knowledge representation and reasoning system.Knowledge representation is the basis of reasoning,and reasoning system is the core part of the automatic problem solving system.Selection and collection of data needed for inductive learning.According to the need,the data of elementary mathematics topics are selected and then used as the training set of inductive learning.After these data are reasoned through the Automatic-Problem-Solving-System,sort out and collect the reasoning process and the results of the problem.The optimization of Automatic-Problem-Solving-System.Although the automatic problem-solving system has completed the initial development,there are still manyshortcomings.In this paper,we will study the existing model of the Automatic-Problem-Solving-System,put forward the optimization scheme,andimprove the automatic problem-solving system according to the optimization scheme.There are two main optimization schemes proposed in this paper.One is to monitor the reasoning process of knowledge base by adding contradictory-downtime rules,find abnormalities in time and give feedback.The other is to combine the idea of mathematical induction with the system so that the system has new reasoning methods and verification strategies.The application of knowledge discovery based on inductive learning in system reasoning.Then sort out and train the collected data,some rules are summed up by machine learning method.Finally,transform these rules into knowledge which is helpful to solve problems and added to the knowledge base of the automatic problem solving system to assist reasoning.After completing the above research and applying the research results to the existing Automatic-Problem-Solving-System,the accuracy and efficiency of the Automatic-Problem-Solving-System have been significantly improved,the accuracy rate has reached 90%,and the average solution time has also been reduced to less than 2 seconds.