Research On Error Model Of Junior High School Mathematics And Its Application In Marking System

With the development of information technology, it has been involved in every aspect of our work and life. In the field of education, the traditional pattern of education means and modes are difficult to satisfy people’s demands, so the concept of intellectual education, internet education were gradually being raised. Thanks to the development of artificial intelligence technology, many experts and scholars began to explore the combination of artificial intelligence and education with the aim to reshape the mode,content, method, system of education. Facing the education revolution, the shortage and uneven distribution of education resources is the most important issue undoubtedly. The main research direction of this article is to seek how the system can judge students answer automatically without teachers participation as well as reference answer,including the right and wrong judgement of each step, and in the wrong circumstances to judge the specific error types. It aims to reduce the human cost of education and point out the weakness in students learning so as to help them to improve their learning, so it is a new exploration for the internet education. The main contents are as follows:Firstly, the research and conginitive modeling of common error types in junior middle school mathematics.Collecting and classifying a large number of junior middle school mathematics common errors, and then analyze and research the phenomenon and research the phenomenon and reasons for these errors in depth with the purpose to build a set of relatively complete system of wrong types, covering almost all types of errors on mathematics concepts in junior high schools. Based on human cognitive processes,analysizing and identifying similarities with the characteristics of each type of error, the research aims to build error cognitive model automatically.Secondly, research and construction of the error model knowledge base.Using NLP(Natural Language Processing) technology, based on the semantic understanding and representation of meaning of the questions and students answer, the research leads to the establishment of inference rules base in line with junior high school math nature, theorems, reasoning and mathematical knowledge as a benchmark.Based on Drools of Forward-Chaining rule engine and symbolic computation tools, the research manages to build an error model knowledge base with known conditions as thestarting point for reasoning. The knowledge base covers all the right answer of students and wrong model related to knowledge inference as much as possible.At last, the application of error model in marking system.Based on error model knowledge base, set up an intelligent marking system which resolve the problem of auto marking cause by problems with more than one solution and diversify of answer. Firstly, the system judge correctness or incorrectness for every step of student’s answer with knowledge base and symbolic computation tools, then using error models to determine the error type for wrong steps, and feedback the specific reasons for the wrong steps to students, so that students understand how and why, achieve precise positioning errors and graders.