基于EKEL的影评挖掘分析方法和系统

Analysis method and system of film's reviews mining based on EKEL

基于EKEL的影评挖掘分析方法是从影评文本中自动获取电影基本知识与评价知识的一种数据挖掘方法.同时该方法是一种结合文法与本体的知识抽取方法,把设计好的影评文法体系与EKEL系统结合,形成一个文本知识挖掘系统,可以有效且准确地抽取影视领域中的基本知识与评价知识.首先设计影评文法,研究设计方法及设计原则,同时结合EKEL系统,对获取到的影评文本语料进行挖掘并分析,产生EKEL分析树,然后对EKEL分析树进行深入分析及后处理.接着进行影评挖掘实验,分析精确率和召回率,并总结本文所研究知识抽取方法的有效性.最后,并对该方法的通用性进行了展望.

Analysis method and system of filmreviews based on EKEL aim to automatically acquire useful film basic knowledge and evaluationknowledge, which is a method of data mining. This method is a text processing method whichcombines with grammar and ontology, the combination of film reviews grammar with EKELsystem form a text knowledge mining system, which can effectivelyand accurately acquire knowledgein the field of film and television. In this paper, we design film reviews grammar, and research on the designing methodandthe de- signingspecification. At the same time, combining EKEL, excavating and analyzing of the acquired film reviews text corpus. Then produce the parse tree of EKEL. Afterwards, elaborately analyze and have a post processing on the parse tree of EKEL. Thenwe proceed to film reviews mining experiment, make a concrete analysis of the pre- cision rate and recall rate. And make a further analysis of the performance of the knowledge extraction methodin this paper＇s study. Finally, we also look ahead to the universality of this method.