互联网汽车品牌口碑挖掘研究

Research on Automobile WOM Mining Based on Domain Ontology

针对获取互联网上汽车评价信息的需求,提出了一种基于领域本体的汽车口碑挖掘方法.该方法在构建汽车领域产品及评论本体的基础上,重点研究了评论句子、比较评论句子的识别及倾向性判定方法.收集了太平洋汽车网的评价语料进行了实验比较,结果验证了所提方法在汽车口碑挖掘应用中的有效性.

Aiming at the requirement of getting automobile judgment on the internet,a method of automobile WOM（Word of Mouth）mining is proposed based on domain ontology.This paper emphatically discusses the identification and tendency judgment of comment sentence and comparative comment sentence on the basis of constructing automobile product and comment ontology.We have collected comment corpus of pacific automobile network to evaluate our proposed methods.The results show that the proposed methods offer more effective performances for mining automobile WOM.