基于FAQ的智能答疑系统中分词模块的设计  

Design of Words Module in Intelligent Q/A System Based on FAQ

作　　者：程节华[1] 

机构地区：[1]安徽科技学院计算机系,安徽蚌埠233100

出　　处：《计算机技术与发展》2008年第7期181-183,186,共4页Computer Technology and Development

基　　金：安徽省自然科学研究项目(KJ2007B159)

摘　　要：在基于FAQ的智能答疑系统中,分词处理是基础和关键。分词质量的好坏直接影响智能答疑系统的准确性。针对实际应用领域的需要,本系统分词模块采取普通词典和专业词典混合的词典设计方案,分词算法采用正向最大匹配的分词算法。为了提高词典的查找速度,分词词典采用Hash表和二维数组的数据结构,根据汉字的内码利用Hash方法,求得在内存的地址,然后计算其索引项的二维数组的下标,对于词典的搜索采用二分查找法。实验结果表明:该分词系统提高了智能答疑系统的效率和准确率。In the intelligent Q/A system based on FAQ, words processing is the basis and key point. Words quality directly influences the intelligence Q/A system accuracy. Aiming at special practical field, a design scheme for words module dictionary which combined with general and special ones is presented. Further more; an algorithm based on forward maximum match is also put into use. In order to enhance the dictionary the search speed, words dictionary uses Hash tables and two - dimension array as data structure, uses the Hash method according to the Chinese encoding to obtain the memory address, and then, calculates the index two - dimensional array sub- script, uses the binary searching method for the dictionary search. The experinaental result indicates that this participle enhances the intelligent Q/A system efficiency and is accurate rate.

关 键 词：自然语言处理 智能答疑 分词 

分 类 号：G434[文化科学—教育学]