Research Of Intelligent Exam System Auxiliary Marking Technology

In recent years,with the rapid development of technology of network and computer replacing the traditional paper test by manual examination is a very active area of research. the course of "Fundamentals of computers " is one of public basic course in most of colleges, after years of transformation of education, Assessment method of the course will be transform automatic group and mark volumes by test systems from the traditional manual way.however, the existing examination system for most of the course just achieve that assessing objective item, the subjective questions related to artificial intelligence and other aspects of technology is still need to solve. The most mark technical of Operation problem on Office were limited to operation of simple text in Word and Excel content, the processing class of complex image is little. Therefore, the study of various kinds of questions for the course of intelligent marking technology has great significance.This thesis on research and implementation of marking the subjective question simulation of artificial marking ideas, natural language processing technology and fuzzy mathematics is defined in close study, proposed fusion algorithm based on key words and phrases and keywords weighted semantic similarity algorithm. extracting the keywords of candidates answer, then calculate the closeness of candidates answer and the standard answer, according to set the weight of keywords, calculation of the final score. The marker on the Office operating questions using VBA technology implement automatically marking, according to the design of solutions and access to Office-related objects, adopting reasonable marking implement and achieving predict function, not only improves the efficiency of the examination time but also reduce the manual grading of objective factors, good or bad feelings of teachers have no effect on the outcome.This thesis is major research the the intelligent marking,part of examination system to, in the course of marked the subjective question reseach on some of the existing reviewers of technology and methods and design process of this scoring system, the first, sloving the candidates answer by clause and participle, and the corresponding semantic analysis, finally, keyword extraction and closeness between the two answers to calculations. After the research on word segmentation, in this paper, the string matching algorithm for maximum matching in the sub-word processing. This article refers to the similarity calculation of fuzzy mathematics and its closeness to improve the definition of the weighted sum in accordance with the method of calculation close. according to the algorithm designed to achieve specific process modules in the smart marking process, for the eventual realization of a "University Computer Foundation" automatic scoring experimental system, the marker for the course achieve to a good marking results.