tailieunhanh - Báo cáo khoa học: "Combining Coherence Models and Machine Translation Evaluation Metrics for Summarization Evaluation"

An ideal summarization system should produce summaries that have high content coverage and linguistic quality. Many state-ofthe-art summarization systems focus on content coverage by extracting content-dense sentences from source articles. A current research focus is to process these sentences so that they read fluently as a whole. | Combining Coherence Models and Machine Translation Evaluation Metrics for Summarization Evaluation Ziheng Lint Chang Liu Hwee Tou Ng and Min-Yen Kant t SAP Research SAP Asia Pte Ltd 30 Pasir Panjang Road Singapore 117440 t Department of Computer Science National University of Singapore 13 Computing Drive Singapore 117417 liuchan1 nght kanmy @ Abstract An ideal summarization system should produce summaries that have high content coverage and linguistic quality. Many state-of-the-art summarization systems focus on content coverage by extracting content-dense sentences from source articles. A current research focus is to process these sentences so that they read fluently as a whole. The current AESOP task encourages research on evaluating summaries on content readability and overall responsiveness. In this work we adapt a machine translation metric to measure content coverage apply an enhanced discourse coherence model to evaluate summary readability and combine both in a trained regression model to evaluate overall responsiveness. The results show significantly improved performance over AESOP 2011 submitted metrics. 1 Introduction Research and development on automatic and manual evaluation of summarization systems have been mainly focused on content coverage Lin and Hovy 2003 Nenkova and Passonneau 2004 Hovy et al. 2006 Zhou et al. 2006 . However users may still find it difficult to read such high-content coverage summaries as they lack fluency. To promote research on automatic evaluation of summary readability the Text Analysis Conference TAC Owczarzak and Dang 2011 introduced a new subtask on readability to its Automatically Evaluating Summaries of Peers AESOP task. 1006 Most of the state-of-the-art summarization systems Ng et al. 2011 Zhang et al. 2011 Conroy et al. 2011 are extraction-based. They extract the most content-dense sentences from source articles. If no post-processing is performed to the generated summaries the .

TỪ KHÓA LIÊN QUAN
crossorigin="anonymous">
Đã phát hiện trình chặn quảng cáo AdBlock
Trang web này phụ thuộc vào doanh thu từ số lần hiển thị quảng cáo để tồn tại. Vui lòng tắt trình chặn quảng cáo của bạn hoặc tạm dừng tính năng chặn quảng cáo cho trang web này.