作者: Fuliang Weng , Yang Liu , Fei Liu
DOI:
关键词:
摘要: User-contributed content is creating a surge on the Internet. A list of "buzzing topics" can effectively monitor and lead people to their topics interest. Yet topic phrase alone, such as "SXSW", rarely present information clearly. In this paper, we propose explore variety text sources for summarizing Twitter topics, including tweets, normalized tweets via dedicated tweet normalization system, web contents linked from well integration different sources. We employ concept-based optimization framework summarization, conduct both automatic human evaluation regarding summary quality. Performance differences are observed input types topics. also provide comprehensive analysis task challenges.