Font Size: a A A

Intelligent Question Answering System In High School Physics Teaching

Posted on:2012-04-13Degree:MasterType:Thesis
Country:ChinaCandidate:C LiFull Text:PDF
GTID:2208330335471230Subject:Modern educational technology
Abstract/Summary:PDF Full Text Request
Network information directly lead to explosive growth--"network information overload". In the vast amount of information resources how to accurately and quickly find answer today, has become the obstacles and the bottle neck in transmission and popularizain of qualified information.To solve the problem of overloaded information, in the 1990s the network information recommendation was brought forward. Today in the education field information recommended popularization has just started. Intelligent question answering system become the education field information recommendation in main forms. Traditional question answering system based on administrator, mostly in online forms, often appears substation timeliness difference, the answer is too limitted. Traditional question answering system is to use keywords to search the metadata. For junior middle school students are not quite up the level of computer languange and skill usually speak natural language.In order to improve the information service quality, this thesis proposes,based on natural language online q&a system to this scheme can not only solve the pbroblem of the elementary and middle school students in processing and extracting keyword deficiency, but also improve the use of network teaching and learning to process knowledge as the guidance in function. Traditional intelligent question answering system is divided into five steps questions pretreatment the Chinese word segmentation, participles to ambiguity, the answer search, result assessment. To improve the traditional information recommended approach, this paper mainly on the traditional algorithm weak link questions pretreatment, the Chinese word segmentation, participles to ambiguity is do the following improvement.Introducing a complete set of questions pretreatment mechanism. Firstly to deal with the Chinese character question the annotation and detachment. Then to give improved maximal matching words segmentation algorithm professional terms and for processing. The accuracy of the DiGaoFen word for word segmentation ambiguity and reduce the probability, introduced appear towards expanding word segmentation algorithm to improve the accuracy of segmentation algorithm.Eliminate parting-words arithmetic ambiguity phenomenon. To eliminate the ambiguity phenomenon word segmentation, the paper introduces maximal matching words segmentation algorithm of improved inverse to tail parting-words arithmetically. Reverse to tail segmentation algorithm introduces, greatly reduces the rate of appearance of the word segmentation ambiguityPerfect result evaluation mechanism. According to the traditional search engines, this paper chooses listing large information and realizes the information accurately. In the information recommended process of selecting new information reliability algorithm and evaluation standard. According to the precision not only sorting and for end users to provide information consultation appraisal for users to refer to them quickly information provided conveniently.In this paper the application of improved algorithm not only can fully assembled the use of network technology in information circulation quickly, conveniently, economcally and other characteristics, but also can achieve exact recommend answer.
Keywords/Search Tags:Chinese characters, positive extender word mark symbols, backward enlarge word algorithm, algorithm, answer evaluation
PDF Full Text Request
Related items