摘要
非正式短文本包含着许多复杂的语义信息,这给文本情感分析研究工作增加了难度,例如不能明确文本所表达的主题、目的和特点。本文提出基于特征向量模型和依存法对非正式短文本作情感分析研究,利用依存句法提取文本的情感元组并计算其情感值,它可以判别文本的情感属性是积极地还是消极地,或者是中立的,并能够通过程度副词判断情感强度。
The short formal text contains many complex semantic information, this increases the difficulty of studying the text sentiment analysis, such as cannot determine what theme, goal or feature the text expressed. We proposed a feature-based vector model and a novel weighting algorithm to study on the short informal text sentiment analysis, and we applies dependency parsing to extract sentiment tuple and calculate the score of sentiment, which can determine the polarity of the text is positive, negative or neutral, and, it can also conclude the sentiment strength by the adverb of degree.
出处
《价值工程》
2015年第23期256-257,共2页
Value Engineering
基金
国家自然科学基金(71471102)
关键词
情感分析
特征向量模型
非正式短文本
sentiment analysis
feature-based vector model
short informal text
