中文摘要：

细颗粒度观点挖掘需要解决观点句识别、要素抽取等关键问题。论文对国内外相关研究进行梳理，归纳现有观点句识别方法并找出影响识别率的原因，分析基于规则和基于统计的两类方法进行要素抽取的优势和不足。研究发现，机器学习方法或人工和机器相结合的方法是要素识别研究的基本趋势，观点本体和观点表达语言规律的研究需要进一步加强，应用领域有待扩展，观点抽取结果挖掘需进一步深入。

英文摘要：

The key steps in fine-grained opinion mining are subjectivity classification and aspect extraction. By reviewing of domestic and foreign related research, the formulated definition of the opinion and aspect is made and the reasons of the low recognition rate are summarized, and then the advantages and disadvantages of two kinds of methods（ based on rules and statistics ） are analyzed. It is found that the semi-supervised machine learning method is a research trend. This paper proposes to strengthen the research on linguistic rules and the ontology of opinion, to expand the application fields and make the comprehensive utilization of the results.