| 摘 要: 微博每天产生的大数据成为市场调研、产品宣传、舆情监控等实际应用高度关注的目标。目前微博要素 级的情感分析缺乏领域针对性，人工处理工作量大。本文提出了基于领域自动分拣的情感要素分析模型，通过实验获 取有价值的博文特征表示，训练评价对象抽取模型和情感倾向性判别模型。本文实现的MSAS(Microblog Sentiment Analysis System)系统能够自动地完成微博数据预处理、情感要素分析和统计分析功能，为相关的应用提供有价值的分 析工具。 |
| 关键词: 情感要素分析；评价对象抽取；MSAS系统 |
|
中图分类号: TP399
文献标识码: A
|
|
| Research and Implementation of the Multi-Domain Sentiment Analysis on Micro-Blog |
|
ZHANG Chao，WANG Longqing
|
(College of Computer Science and Technology, Donghua University, Shanghai 201620, China )
|
| Abstract: Big data generated on Micro-blog has caught the attention of various aspects of practical application,such as market survey,product promotion,public opinion monitoring,etc.The current sentiment analysis method is severely lacking in pertinence to specific domains,and the pre-processing work is labor intensive.The paper proposes a feature-level sentiment analysis model based on automated domain data filtering,and conducts experiments to achieve the valuable feature presentation of micro-blog articles and train the opinion target extraction model and the sentimental classification model. The MSAS(Micro-blog Sentiment Analysis System)constructed in this study can automatically implement the pre-processing of micro-blog raw data,the analysis of sentimental factors and the statistics and analysis function,offering valuable analysis tools for relevant application. |
| Keywords: sentiment factor analysis;opinion target extraction;MSAS |
