基于文本分析的知识获取系统设计与实现  

Design and Implementation of Text Analysis Based Knowledge Acquisition System

作　　者：姚金国[1] 代志龙[1] 

机构地区：[1]复旦大学计算机科学与技术学院,上海200433

出　　处：《计算机工程》2011年第2期157-159,共3页Computer Engineering

基　　金：国家自然科学基金资助项目(60303007)

摘　　要：知识获取一直以来都是构建专家系统的瓶颈问题。针对该问题,利用自然语言处理技术,设计并实现一个针对化学科技文献进行分析的知识获取系统,并对其关键技术进行分析。系统对输入文本进行分词及词性标注,在此基础上,使用Tregex在句法分析树上进行实体识别,同时利用依存关系进行搭配词识别。实验结果表明了该系统的有效性。Knowledge acquisition is a bottleneck to develop expert system. Aiming at this problem, it designs and implements a knowledge acquisition system to analyze the chemical literatures. Meanwhile, key techniques are expatiated. NLP approaches are used in this system to facilitate the automated extraction of chemical knowledge. The system segments the input text into words and assigns part of speech to each word. Tregex is used to identify the named entities on parsing trees and typed dependencies are utilized to extract qualifier for each entity. Experimental results indicate that the approaches for this task are highly efficient.

关 键 词：知识获取 句法分析 文本分析 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]