论文信息 - CLASSY 2007 at DUC 2007

CLASSY 2007 at DUC 2007

The IDA/CCS summarization system, CLASSY (Clustering, Linguistics, And Statistics for Summarization Yield), was enhanced in several areas for this year’s DUC. Our sentence splitting and trimming algorithms continue to be improved. Signature terms were improved by using the AQUAINT data as the background. Redundancy removal was also considerably improved employing LSI and a new variant of QR. We proposed a new way to determine paragraph breaks. In addition, a sub-cluster redundancy removal method was developed to tackle the update summary task. We summarize our results and analyze the relationship between ROUGE scores and responsiveness.

John M. Conroy | Dianne P. O'Leary | Judith D. Schlesinger

[1] Martin F. Porter,et al. An algorithm for suffix stripping , 1997, Program.

[2] Ted Dunning,et al. Accurate Methods for the Statistics of Surprise and Coincidence , 1993, CL.

[3] John Doull. Back to Basics , 1997 .

[4] Ani Nenkova,et al. The Impact of Frequency on Summarization , 2005 .

[5] Dianne P. O'Leary,et al. Topic-Focused Multi-Document Summarization Using an Approximate Oracle Score , 2006, ACL.

[6] John M. Conroy,et al. CLASSY Arabic and English Multi-Document Summarization , 2006 .

[7] John M. Conroy,et al. Three CLASSY ways to Perform Arabic and English Multi-Document Summarization , 2005 .

[8] John M. Conroy,et al. Back to Basics: CLASSY 2006 , 2006 .

[9] Mirella Lapata,et al. Automatic Evaluation of Text Coherence: Models and Representations , 2005, IJCAI.

[10] Eduard H. Hovy,et al. The Automated Acquisition of Topic Signatures for Text Summarization , 2000, COLING.