RI: Small: Collaborative Research: Word Sense and Multilingual Subjectivity Analysis

RI：小型：协作研究：词义和多语言主观性分析

• 批准号: 0916046
• 负责人: Janyce Wiebe
• 金额: $ 22.5万
• 依托单位: University of Pittsburgh
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2009
• 资助国家: 美国
• 起止时间: 2009-09-01 至 2013-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0916046&HistoricalAwards=false
• 关键词: RI; Small; Collaborative; Research; Word

Approaches to subjectivity and sentiment analysis often rely onmanually or automatically constructed lexicons. Most such lexicons arecompiled as lists of words, rather than word meanings ("senses").However, many words have both subjective and objective senses as wellas senses of different polarities, which is a major source ofambiguity in subjectivity and sentiment analysis. The proposed workaddresses this gap, by investigating novel methods for subjectivitysense labeling, and exploiting the results in sense-aware subjectivityand sentiment analysis. To achieve these goals, three researchobjectives are targeted. The first is developing methods for assigningsubjectivity labels to word senses in a taxonomy. The second isdeveloping contextual subjectivity disambiguation techniques toeffectively make use of the word sense subjectivity annotations. Thethird is applying these techniques to multiple languages, includinglanguages with fewer resources than English. The project will havebroader impacts in both research and education. First, it will makesubjectivity and sentiment resources and tools more widely available,in multiple languages, to the research community, which will helpadvance the state of the art in automatic subjectivity analysis, whichin turn will benefit end applications. Second, several educationalgoals will be pursued: training graduate and undergraduate students incomputational linguistics; augmenting artificial intelligence courseswith projects based on the proposed research, which will offerstudents hands-on experience with natural language processingresearch; and reaching out to women and minorities to increase theirexposure to text processing technologies and access to researchopportunities.

主观性和情感分析的方法通常依赖于手动或自动构建的词典。然而，许多词汇同时具有主观意义和客观意义以及不同极性的意义，这是主观性和情感分析中歧义的主要来源。 拟议的工作解决了这一差距，通过调查新的方法为subjectivitysense标签，并利用感官感知的subjectivityand情感分析的结果。 为了实现这些目标，有三个研究目标。第一个是开发方法，在分类法中为词义分配主观标签。二是发展语境主观性消歧技术，有效利用词义主观性标注.第三是将这些技术应用于多种语言，包括资源比英语少的语言。 该项目将在研究和教育方面产生更广泛的影响。 首先，它将使主观性和情感资源和工具以多种语言更广泛地提供给研究社区，这将有助于推进自动主观性分析的最新技术水平，这反过来又将有利于最终应用。 第二，将追求几个教育目标：培训研究生和本科生学习计算语言学;通过基于拟议研究的项目增加人工智能课程，这将为学生提供自然语言处理研究的实践经验;并接触女性和少数民族，以增加他们对文本处理技术的兴趣和获得研究机会的机会。