Online Automatic Question Answering System Based On Chinese Word Segmentation Technology

With the development of Internet and network technology, now computer technologyhas been applied in many fields. Online teaching platform is one of the typical andimportant branches. Answering system is closely related to online teaching platform. It isan indispensable system of them. It can timely answer students’ questions, disabuse theiranxieties and so on. Automatic answering system depends on the development of integratedapplication of various techniques. Among them, Chinese word segmentation technology isone of the key links to it. It can be said, without it, the answering system cannot workeffectively.This article is based on Chinese word segmentation technology, is trying to design anon-line automatic question answering system. The first part is literature review. And wediscuss the background of this technology as well as the current researches; the second partis a brief introduction of the Chinese word segmentation technology; the last part isconcerned on the practice of this technology. Through the word segmentation on thestudents’ questions， answers are matched in the database, the matching degree of higherin the database of question and answer feedback to students, to achieve the intelligentquestion answering system. The connotation of the algorithm is as follows: Algorithm basedon segmentation fields all segmentation pretreatment, get all possible segmentation forms,and then based on the segmentation of different structure form adopts different processingstrategies for ambiguity processing. Meanwhile, it is hoped that can be achieved on thedictionary a visit for many times to use, due to ease of frequent access dictionary efficiencyproblem. Also helps to solve the problem of word segmentation in stages.According to the characteristics of the answering system，Study of this code，Segmentation method based on the full segmentation graph algorithmse mbedded in the Lucene full-text retrieval system. According to the Lucene search engine has astrong object-oriented features, getting the central part of the abstract class and open classof msnm-lib package together, then from time to time online automatic function ofAnswering system achieved. Lucene system uses a fuzzy query, find the answer in thedatabase with the keywords and feedback to students.This paper summarizes the whole work, also to the further research direction wasanalyzed and the prospect.