关键词: 观点分析 观点自动检测 观点持有者 SVM 距离加权计算

Abstract:

This paper presents an approach based on SVM and distance-weighted computing to automatically analyze opinions, which includes two aspects： opinion automatic detection and opinion holder automatic extraction. We establish a series of related resources for opinion analysis, such as the opinion-bearing verb set, etc. Based on these related resources, we accomplish the detection of opinions via SVM. On the basis of opinion automatic detection, we use t he opinion-bearing verb as a key indicator, apply a distance-weighted computing method to extract the opinion holder （or holders） of a given sentence,   , and expand the opinion holders to noun phrases by pattern matching. Thus, the results of opinion holder automatic extraction are further improved.

Key words: opinion analysis, opinior, automatic detection, opinion holder, SVM, distance-weighted computing