RI: Multi-Faceted Comparative Text Summarization

文本摘要生成文档的简明摘要以帮助用户快速地消化信息，并且由于信息的爆炸式增长而变得越来越重要。现有的大多数摘要方法只能生成一个简单的句子列表的非结构化摘要。然而，在许多应用程序中，我们希望生成一个（更结构化的）多方面的比较摘要，其中句子被分组为多个方面，并在不同的视图中进行比较。例如，关于笔记本电脑的意见摘要可能会将句子分为“电池寿命”和“记忆”等方面，并在每个方面分别列出积极和消极的意见。本项目旨在系统地研究这一新的文摘问题（称为多面比较文摘）。PI将开发通用的概率方法，可以应用于不同领域的问题的多个实例。其基本思想是利用概率混合模型对一组待摘要的文本文档进行建模，提取文本文档中的多个刻面和每个刻面的多个视图，然后利用提取的刻面和视图生成刻面标签，并为不同的刻面-视图组合选择句子.本文的研究为文本摘要开辟了一个新的研究方向。所开发的方法可以直接应用于为所有人提供更有用的Web上的产品意见检索服务，帮助多个领域的科学家更有效地消化科学文献，从而加速科学发现，所有的研究资源和成果都将通过出版物和可下载的软件和数据传播给研究界。研究结果也将加强目前的信息检索课程，从而改善信息技术劳动力的教育。

Text summarization generates a concise summary of documents tohelp a user quickly digest information and has become increasingly important due to the explosive growth of information. Most existing summarization methods can only generate an unstructured summary with a simple list of sentences. In many applications, however, we want to generate a (more structured)multi-faceted comparative summary, in which sentences are grouped into multiple facets and compared across different views. For example, a summary about laptop opinions may group sentences into facets such as ``battery life'' and ``memory'' and separate sentenceswith positive and negative opinions in each facet. This project aims to systematically study this new summarization problem (called multi-facetedcomparative summarization). The PI will develop general probabilistic approaches that can be applied to multiple instances of the problem in different domains. The basic idea is to use probabilistic mixture models to model and extractthe multiple facets and multiple views of each facet in a set of text documents to be summarized.The extracted facets and views are then used to generate facet labels and select sentences fordifferent facet-view combinations. The proposed research will open up a new research direction in text summarization. The developed methods can be directly applied to provide more usefulproduct opinion retrieval service on the Web to all people and help scientists in multiple fieldsto digest scientific literature more effectively, thus speeding up scientific discovery.All research resources and results will be disseminated to the research community through publications and downloadable software and data. The research results will alsoenhance the current curriculum in information retrieval, leading to improved education of the information technology workforce.