K2Q: Generating Natural Language Questions from Keywords with User Refinements

Proceedings of the 5th International Joint Conference on Natural Language Processing, pages 947–955, Chiang Mai, Thailand, November 8 – 13, 2011. c⃝2011 AFNLP K2Q: Generating Natural Language Questions from Keywords with User Refinements Zhicheng Zheng Tsinghua University zhengzc04@gmail.com Xiance Si Google Inc. sxc@google.com Edward Y. Chang Google Inc. edchang@google.com Xiaoyan Zhu Tsinghua University zxy-dcs@tsinghua.edu.cn Abstract Garbage in and garbage out. A Q&A sys- tem must receive a well formulated ques- tion that matches the user’s intent or she has no chance to receive satisfactory an- swers. In this paper, we propose a key- words to questions (K2Q) system to assist a user to articulate and refine questions. K2Q generates candidate questions and re- finement words from a set of input key- words. After specifying some initial key- words, a user receives a list of candidate questions as well as a list of refinement words. The user can then select a satisfac- tory question, or select a refinement word to generate a new list of candidate ques- tions and refinement words. We propose a User Inquiry Intent (UII) model to de- scribe the joint generation process of key- words and questions for ranking questions, suggesting refinement words, and generat- ing questions that may not have previously appeared. Empirical study shows UII to be useful and effective for the K2Q task. 1 Introduction Keyword search has long been considered an un- natural undertaking task, but it works well with search engines. When entering a question to a Q&A system such as Yahoo! Answers and Quora, however, the keyword paradigm simply does not work. A question must be articulated specifically in the form of natural language. For instance, key- words such as New York restaurant is ambiguous in its intent. The user may want restaurants in New York, or want to know the tipping practice at New York restaurants, or something else under the myr- iad other possible interpretations. For a question to be answered, the asker must articulate her in- tent with sufficient specificity. Partly because we have been detrained by search engine from writ- ing a question in a complete sentence, and partly because a question is hard to articulate completely when the asker is in a learning mode, a Q&A sys- tem should provide tools to users to help them clar- ify their questions so as to get good answers. In this paper, we propose K2Q, a system that converts keywords to questions by considering both query history and user feedback. More specifically, given a set of keywords, K2Q gen- erates a list of ranked questions as well a list of refinement words. A user can select a satisfactory question, or she can select a refinement word to generate a new list of candidate questions and re- finement words. This process iterates until the user finds a good question matching her intent or quits. To build an effective K2Q system, there are three aspects that need to be researched. A user issues a question typically because the desired in- formation …

Content truncated to 500 words