曝光台 注意防骗
网曝天猫店富美金盛家居专营店坑蒙拐骗欺诈消费者
for Computational Linguistics on Computational Linguistics,
pages 25–32.
Chin-Yew Lin and Eduard Hovy. 2000. The automated acquisition
of topic signatures for text summarization. In Proceedings
of the 18th conference on Computational linguistics,
pages 495–501.
Chin-Yew Lin and Eduard Hovy. 2003. Automatic evaluation
of summaries using n-gram co-occurance statistics. In Proceedings
of HLT-NAACL 2003.
Chin-Yew Lin, Guihong Cao, Jianfeng Gao, and Jian-Yun Nie.
2006. An information-theoretic approach to automatic evaluation
of summaries. In Proceedings of the Human Language
Technology Conference of the NAACL, Main Conference,
pages 463–470.
Chin-Yew Lin. 2004. ROUGE: a package for automatic evaluation
of summaries. In ACL Text Summarization Workshop.
Inderjeet Mani, Gary Klein, David House, Lynette Hirschman,
and Therese Firmin abd Beth Sundheim. 2002. Summac: a
text summarization evaluation. Natural Language Engineering,
8(1):43–68.
Ryan McDonald. 2007. A study of global inference algorithms
in multi-document summarization. In ECIR, pages 557–564.
Kathleen McKeown, Regina Barzilay, David Evans, Vasileios
Hatzivassiloglou, Barry Schiffman, and Simone Teufel.
2001. Columbia multi-document summarization: Approach
and evaluation. In DUC’01.
comparison with update inputs only avg. comparisons with update and background
features pyramid score responsiveness pyramid score responsiveness
JSD divergence -0.827 -0.764 -0.716 -0.669
JSD divergence smoothed -0.825 -0.764 -0.713 -0.670
% of ip topic wds in summ 0.770 0.709 0.677 0.616
KL divergence summ-inp -0.749 -0.709 -0.651 -0.624
KL divergence inp-summ -0.741 -0.717 -0.644 -0.638
cosine inp-summ 0.727 0.691 0.649 0.631
% of summary = topic wd 0.721 0.707 0.647 0.636
topic overlap inp- summ 0.707 0.674 0.645 0.619
multinomial summ prob 0.284 0.355 0.152 0.224
unigram summ prob -0.093 0.038 -0.151 -0.053
regression 0.789 0.605 0.699 0.522
regression combining features comparing with background and update inputs (without averaging)
correlations = 0.8058 with pyramid 1, 0.6729 with responsiveness
pyramid score responsiveness
features max min sig %sig a0.5 %a0.5 max min sig %sig a0.5 %a0.5
JSD smoothed -0.753 -0.269 41 85.4 23 47.9 -0.747 -0.266 36 75.0 16 33.3
JSD -0.746 -0.291 41 85.4 22 45.8 -0.738 -0.263 36 75.0 16 33.3
KL summ-inp -0.739 -0.293 41 85.4 20 41.7 -0.705 -0.275 37 77.1 15 31.3
% of sign from inp 0.778 0.277 38 79.2 17 35.4 0.706 0.297 29 60.4 13 27.1
cosine overlap 0.665 0.275 33 68.8 10 20.8 0.685 0.267 28 14.6 7 14.6
% summ sign terms 0.737 0.263 32 66.7 11 22.9 0.672 0.265 28 58.3 6 12.5
topic overlap 0.679 0.264 31 64.6 9 18.8 0.665 0.274 26 54.2 5 10.4
KL inp-summ -0.663 -0.281 30 62.5 9 18.8 -0.600 -0.285 24 50.0 5 10.4
mult. summ prob 0.479 0.267 12 25.0 0 0.0 0.547 0.262 13 27.1 1 2.1
uni. summ prob 0.363 0.362 1 2.1 0 0.0 0.266 0.266 1 2.1 0 0.0
regression 0.765 0.284 40 83.3 19 39.6 0.659 0.285 29 60.4 10 20.8
ROUGE-1 recall 0.842 0.392 48 100 41 85.4 0.811 0.268 46 95.8 30 62.5
ROUGE-2 recall 0.913 0.355 47 97.9 39 81.3 0.816 0.286 47 97.9 28 58.3
Table 4: Spearman correlations between fully automatic evaluation and manually assigned system scores for update summarization.
Results are reported separately for features comparing update summaries with the update input only or with both update and
background inputs and averaging the two (macro level). At the per-input level, only results for features comparing with update
inputs are reported.
Kathleen McKeown, Rebecca Passonneau, David Elson, Ani
Nenkova, and Julia Hirschberg. 2005. Do summaries help?
a task-based evaluation of multi-document summarization.
In SIGIR.
Andrew H. Morris, George M. Kasper, and Dennis A. Adams.
1992. The effects and limitations of automatic text condensing
on reading comprehension. Information System Research,
3(1):17–35.
Ani Nenkova and Rebecca Passonneau. 2004. Evaluating content
selection in summarization: The pyramid method. In
HLT/NAACL.
Ani Nenkova, Rebecca Passonneau, and Kathleen McKeown.
2007. The pyramid method: Incorporating human content
selection variation in summarization evaluation. ACMTrans.
Speech Lang. Process., 4(2):4.
Paul Over, Hoa Dang, and Donna Harman. 2007. Duc in context.
Inf. Process. Manage., 43(6):1506–1520.
Dragomir Radev and Daniel Tam. 2003. Single-document and
multi-document summary evaluation via relative utility. In
Poster session, International Conference on Information and
Knowledge Management (CIKM’03).
Dragomir Radev, Simone Teufel, Horacio Saggion, Wai Lam,
中国航空网 www.aero.cn
航空翻译 www.aviation.cn
本文链接地址：
航空资料36(47)
