Answering opinion questions from blogs

如果今晚美国总统辞职，仅仅几分钟后，各种观点、信仰和其他评论就会充斥网络。随着越来越多的文本信息在网上可用，开发自然语言工具来帮助快速分析非正式文本以及时做出决策变得很有必要。问答（QA）的最新发展使得用户可以用自然语言提出问题并获得特定的答案，而不是整个文档。然而，到目前为止，QA中的大多数工作都涉及回答来自正式和信息丰富来源（如报纸文章）的基于事实的问题。然而，随着越来越多的人使用网络来表达自己的意见，回答非正式文件中的意见问题将成为一个重要的问题。

If tonight, the president of the U.S. were to resign, only a few minutes later, opinions, beliefs and other comments would already flood the Web. As more and more textual information is available online, it has become necessary to develop natural language tools to help analyze informal texts quickly to make timely decisions. Recent developments in Question Answering (QA) have made it possible for users to ask a question in natural language and receive a specific answer rather than an entire document. However, to date, most work in QA has been involved in answering fact-based questions from formal and information-rich sources such as newspaper articles. However, as more and more people use the Web to express their opinions, the problem of answering opinion questions from informal documents will become an important issue.