The Research And Application Of The Key Technology In The Online Examination System

In the information technology continues to progress today, more and more disciplines of information technology into daily teaching process, making the process a lot of the teaching curriculum reform, one of the biggest innovations is the introduction of paperless online examination system. Earlier online examination system has a strong discipline targeted, common in computer-related subjects course examination, which objective questions, such as single choice, multiple choice, mostly, subjective questions prepared places fixed code format or computer the main steps. For some liberal arts courses, on-line examination system application only limited to objective questions, subjective questions rarely appear in the online exam.The main problem for the judge’s subjective questions in the process of automatic marking the presence of the following:1, examine the contents of more subjective questions for students to play the main exam is difficult to give a fixed answer, in the conventional multi-hand scoring is also treated as a knowledge-point scoring. The subject of the existing technology and word segmentation algorithm for analysis, comparing the higher efficiency segmentation techniques.2, subjective questions of knowledge covered many and complex, relative to other Chinese language, sentences are complicated, large vocabulary, and a large number of synonyms, synonyms, the same words have different meanings in different contexts which,in automatic marking process, the knowledge point for effective sorting is often the largest amount of work a job. The topic of the word present in the system disambiguation depth analysis and identification ambiguity and disambiguation both do a little improvement,bidirectional scanning + Full segmentation algorithm integrated approach to improve the accuracy of segmentation.3, subjective questions during the automatic marking process, due to the high degree of complexity of the Chinese language, the answer to how to improve the degree of similarity with the standard answer of the candidates is a comparison of the accuracy of the extremely difficult problem. This paper uses the similarity calculation method based on dependency grammar and standard answers for candidates answer compare calculations..