Font Size: a A A

Research On The Technology Of Problem Comprehension For Automatic Problem Solving Of Geography Course

Posted on:2018-02-15Degree:MasterType:Thesis
Country:ChinaCandidate:L Y ZhuFull Text:PDF
GTID:2348330518984084Subject:Computer technology
Abstract/Summary:PDF Full Text Request
As one of the research hotspot in the field of artificial intelligence and natural language processing,automatic problem solving focuses on solving problems of some courses by computer.The problem comprehension technology is to let the computer automatically understand the meaning of the text,through the text of the lexical analysis,syntax analysis and semantic understanding and other related theoretical research and study,so as to achieve the computer automatically solve the problem.The current automatic problem solving research is mainly for mathematics courses,for other courses are less research.The geography curriculum as a liberal arts subject,but contains a wealth of knowledge,"on the astronomical,under the geography" also illustrates the importance of geography curriculum.This paper mainly studies the problem comprehension technology of automatic problem solving,and its research contents are as follows:(1)In this paper,the SVM algorithm is applied to the classification of geography questions,and the learning function of the test text is trained by using the Linear kernel function in the LIBSVM classification package.The TFIDF extract keywords,generate feature vector,build classification model for testing The experimental results show that the classification function of the linear kernel function is better and the classification accuracy is more than 80%.The application of the text classification algorithm in the field of geography curriculum is realized.(2)Aiming to automatic problem solving of geography course,the semantic relationship between the conceptual entities is further obtained by semantic analysis of the problems and the options on the basis of constructing the ontology of geographical problems,and then the semantic relation is transformed into the problem-solving rules.The experimental results on the collected geographic problem sets show that the acquired rules can obviously improve the solution of the text-type problems,and also have some auxiliary effects on the solution of the chart-type problems.
Keywords/Search Tags:SVM algorithm, automatic problem-solving, geographical domain ontology, problem-solving rules
PDF Full Text Request
Related items