CatSumm: Spektral Çizge Bölmeleme ve Düğüm Merkeziliklerine Dayalı Çıkarıcı Metin Özetleme

Abstract

Bu çalışmada, çok belgeli metin özetleme için yeni bir yöntemi CatSumm'ı (Cengiz, Ali, Taner Özetleme) tanıtıyoruz. Önerilen yöntem, üç ana adıma göre bir özet oluşturmaktadır: Giriş metinlerinin temsili, CatSumm modelinin ana aşamaları ve cümle puanlama. Girilen metinlerin gösterimi aşamasında kelime grupları arasındaki anlamsal bağlılığı korumak için bir Metin İşleme yazılımı tanıtılmış ve kullanılmıştır. CatSumm modelinin ana aşamalarından biri olan Spektral Cümle Kümeleme (SCK), spektral çizge bölmeleme sonrasında elde edilen alt çizgelerin oransal değerlerinden elde edilen özetleme işlemidir. Standart sapma ile hesaplanan bir eşik değerinin altında kalan cümlelerin özete dahil edilemeyeceği varsayımıyla, yöntemin ana aşamalarından bir diğeri de süper kenarların elde edilmesidir. Son olarak, araştırma kapsamında metin özetleme amacıyla CatSumm yönteminin sonucu, Belge Anlama Konferansı (DUC-2004, DUC-2002) veri setleri üzerinde ROUGE metrikleri ile ölçülmüştür. Yapılan ölçümler sonucunda Catsumm'ın benzer yöntemler kullanan özetleme yaklaşımlarından daha iyi performans gösterdiği tespit edilmiştir.

Keywords

• Belge Özetleme
• Özetleme
• Çıkarıcı özetleme
• Spektral bölmeleme
• Çizge tabanlı özetleme
• Kenar azaltma

CatSumm: Extractive Text Summarization based on Spectral Graph Partitioning and Node Centrality

Abstract

In this paper, we introduce CatSumm (Cengiz, Ali, Taner Summarization), a novel method for multi-document document summarisation. The suggested method forms a summarization according to three main steps: Representation of input texts, the main stages of the CatSumm model, and sentence scoring. A Text Processing software, is introduced and used to protect the semantic loyalty between word groups at stage of representation of input texts. Spectral Sentence Clustering (SSC), one of the main stages of the CatSumm model, is the summarization process obtained from the proportional values of the sub graphs obtained after spectral graph segmentation. Obtaining super edges is another of the main stages of the method, with the assumption that sentences with weak values below a threshold value calculated by the standard deviation (SD) cannot be included in the summary. Using the different node centrality methods of the CatSumm approach, it forms the sentence rating phase of the recommended summarising approach, determining the significant nodes and hence significant nodes. Finally, the result of the CatSumm method for the purpose of text summarisation within the in the research was measured ROUGE metrics on the Document Understanding Conference (DUC-2004, DUC-2002) datasets. As a result of the measurements performed, it was determined that the Catsumm performs better than known summarization approach.

Keywords

• Document Summarization
• Summarization
• Extractive summarization
• Spectral partitioning
• Graph-based summarization
• Edge Reduction

References

1. [1] Durmaz O., 2011. Metin Sınıflandırmada Boyut Azaltmanın Etkisi ve Özellik Seçimi. doi: 10.1360/zd-2013-43-6-1064.
2. [2] Hark C., Uçkan T., and Seyyarer A., Karci A., 2018. Metin Özetleme İçin Çizge Tabanlı Bir Öneri. in IDAP 2018 - International Artificial Intelligence and Data Processing Symposium, pp. 1–6.
3. [3] Ş. Canberk, G. , Sağıroğlu, 2006. Bilgi ve Bilgisayar Güvenliği : Casus Yazılımlar ve Korunma Yöntemleri. Ankara: Grafiker Yayıncılık,.
4. [4] T. UÇKAN, C. HARK, E. SEYYARER, and A. KARCI, 2019. Ağırlıklandırılmış Çizgelerde Tf-Idf ve Eigen Ayrışımı Kullanarak Metin Sınıflandırma, Bitlis Eren Üniversitesi Fen Bilim. Derg., doi: 10.17798/bitlisfen.531221.
5. [5] C. Hark, T. Uckan, E. Seyyarer, and A. Karci, 2019. Extractive Text Summarization via Graph Entropy Çizge Entropi ile Çikarici Metin Özetleme. doi: 10.1109/IDAP.2019.8875936.
6. [6] C. Hark, A. Seyyarer, T. Uçkan, and A. Karci, 2017. Doǧal dil işleme yaklaşimlari ile yapisal olmayan dökümanlarin benzerliǧi. - International Artificial Intelligence and Data Processing Symposium, pp. 1–6, doi: 10.1109/IDAP.2017.8090306.
7. [7] D. R. Radev, E. Hovy, and K. McKeown, 2002. Introduction to the special issue on summarization. Comput. Linguist., vol. 28, no. 4, pp. 399–408.
8. [8] G. Erkan and D. R. Radev, 2004. Lexrank: Graph-based lexical centrality as salience in text summarization. J. Artif. Intell. Res., vol. 22, pp. 457–479.

9. [9] D. Das and A. F. T. Martins, 2007. A survey on automatic text summarization. Lit. Surv. Lang. Stat. II course C., vol. 4, no. 192–195, p. 57.
10. [10] O. Kaynar, Y. Görmez, Y. E. Işık, and F. Demirkoparan,2017. Comparison of graph based document summarization method, in 2017 International Conference on Computer Science and Engineering (UBMK), pp. 598–603, doi: 10.1109/UBMK.2017.8093475.
11. [11] M. Kutlu, C. Cigir, and I. Cicekli,2010. Generic text summarization for Turkish. Comput. J., vol. 53, no. 8, pp. 1315–1323.
12. [12] R. M. Alguliev, R. M. Aliguliyev, and M. S. Hajirahimova, 2012. GenDocSum+ MCLR: Generic document summarization based on maximum coverage and less redundancy, Expert Syst. Appl., vol. 39, no. 16, pp. 12460–12473.
13. [13] V. Dalal and L. Malik, Dec. 2013.A Survey of Extractive and Abstractive Text Summarization Techniques, in 2013 6th International Conference on Emerging Trends in Engineering and Technology, pp. 109–110, doi: 10.1109/ICETET.2013.31.
14. [14] C. HARK, T. UÇKAN, E. SEYYARER, and A. KARCI,2019. Metin Özetlemesi için Düğüm Merkezliklerine Dayalı Denetimsiz Bir Yaklaşım. Bitlis Eren Üniversitesi Fen Bilim. Derg., doi: 10.17798/bitlisfen.568883.
15. [15] R. Mihalcea and P. Tarau,2005. A Language Independent Algorithm for Single and Multiple Document Summarization. in Proceedings of IJCNLP 2005, 2nd International Join Conference on Natural Language Processing, pp. 19–24.
16. [16] K. Sarkar, K. Saraf, and A. Ghosh, 2015. Improving graph based multidocument text summarization using an enhanced sentence similarity measure. in 2015 IEEE 2nd International Conference on Recent Trends in Information Systems, ReTIS 2015 - Proceedings, pp. 359–365, doi: 10.1109/ReTIS.2015.7232905.
17. [17] A. Joshi, E. Fidalgo, E. Alegre, and L. Fernández-Robles,2019. SummCoder: An unsupervised framework for extractive text summarization based on deep auto-encoders. Expert Syst. Appl., vol. 129, pp. 200–215 doi: 10.1016/j.eswa.2019.03.045.
18. [18] R. Mihalcea and P. Tarau, 2004.TextRank: Bringing Order into Texts. in Proceedings of the ACL 2004 on Interactive poster and demonstration sessions -, vol. 85, pp. 20-es, doi: 10.3115/1219044.1219064.
19. [19] D. Parveen, H.-M. Ramsl, and M. Strube,2015. Topical coherence for graph-based extractive summarization. in Proceedings of the 2015 Conference on Empirical Methods in Natural Language Processing, pp. 1949–1954.
20. [20] C. Hark and A. Karcı, 2020. Karcı summarization: A simple and effective approach for automatic text summarization using Karcı entropy. Information processing & management, vol. 57, no. 3, p. 102187, doi: 10.1016/j.ipm.2019.102187.
21. [21] T. Uçkan and A. Karcı,2019. Extractive multi-document text summarization based on graph independent sets. doi: 10.1016/j.eij.2019.12.002.
22. [22] H. P. Luhn 1958. The Automatic Creation of Literature Abstracts. IBM J. Res. Dev., vol. 2, no. 2, pp. 159–165, Apr. 1958, doi: 10.1147/rd.22.0159.
23. [23] H. P. Edmundson,1969. New methods in automatic extracting. J. ACM, vol. 16, no. 2, pp. 264–285.
24. [24] C. Mallick, A. K. Das, M. Dutta, A. K. Das, and A. Sarkar, 2019. Graph-Based Text Summarization Using Modified TextRank. in Soft Computing in Data Analytics, Springer, pp. 137–146.
25. [25] S. Pouriyeh, M. Allahyari, Q. Liu, H. R. Arabnia, Y. Qu, and K. Kochut, “Graph-based Ontology Summarization: A Survey.”
26. [26] M. Allahyari et al.,2017. Text Summarization Techniques: A Brief Survey. doi: 10.1145/nnnnnnn.nnnnnnn.
27. [27] M. Nasr Azadani, N. Ghadiri, and E. Davoodijam,2018. Graph-based biomedical text summarization: An itemset mining and sentence clustering approach. J. Biomed. Inform., vol. 84, pp. 42–58, doi: 10.1016/J.JBI.2018.06.005.
28. [28] J. Dhondt , P.-A. Verhaegen, J. Vertommen, D. Cattrysse, and J. R. Duflou,2011.Topic identification based on document coherence and spectral analysis. doi: 10.1016/j.ins.2011.04.044.
29. [29] T. UÇKAN, C. HARK, and A. KARCİ, 2020. SSC: Clustering Of Turkish Texts By Spectral Graph Partitioning. J. Polytech., doi: 10.2339/politeknik.684558.
30. [30] A. Karci, “Çizge Algoritmaları ve Çizge Bölmeleme 2007. Fırat Universitesi, 1998.
31. [31] U. Von Luxburg,2018. A Tutorial on Spectral Clustering. Accessed: [Online]. Available: www.springer.com.
32. [32] B. Slininger,2018. Fiedler’s Theory of Spectral Graph Partitioning. Accessed: Dec. 20, 2018. [Online]. Available: http://www.cs.berkeley.edu/~demmel/.
33. [33] Niles Robert, “Statistics: Definition of Standard Deviation.” .
34. [34] A. Bavelas,1948. A Mathematical Model for Group Structures. Human Organization, vol. 7, no. 3. pp. 16–30, doi: 10.17730/humo.7.3.f4033344851gl053.
35. [35] M. A. Fattah and F. Ren, 2009.GA, MR, FFNN, PNN and GMM based models for automatic text summarization. Comput. Speech Lang., vol. 23, no. 1, pp. 126–144, doi: 10.1016/j.csl.2008.04.002.
36. [36] F. Boudin et al., 2013. A Comparison of Centrality Measures for Graph-Based Keyphrase Extraction To cite this version : HAL Id : hal-00850187 A Comparison of Centrality Measures for Graph-Based Keyphrase Extraction.
37. [37] Alex Kosorukoff,2011. Social Network Analysis Theory and Applications. Passmore, D. L,.
38. [38] M. R. Garey and D. S. Johnson, 1979 Computers and intractability : a guide to the theory of NP-completeness. W.H. Freeman.
39. [39] M. McPherson, L. Smith-Lovin, and J. M. Cook, 2001. Birds of a Feather: Homophily in Social Networks,” Annu. Rev. Sociol., vol. 27, no. 1, pp. 415–444, doi: 10.1146/annurev.soc.27.1.415.
40. [40] B. N. Analysis,2016. Centrality and Hubs. no. 1979, doi: 10.1016/B978-0-12-407908-3.00005-4.
41. [41] NIST, “Document Understanding Conferences,” NIST. .
42. [42] C. Y. Lin, 2004 Rouge: A package for automatic evaluation of summaries. Proc. Work. text Summ. branches out (WAS 2004), pp. 25–26, doi: 10.1.1.111.9426.
43. [43] C.-Y. Lin and E. Hovy, “Automatic Evaluation of Summaries Using N-gram Co-Occurrence Statistics.” Accessed: May 01, 2019. [Online]. Available: https://www.aclweb.org/anthology/N03-1020.
44. [44] S. Xiong and D. Ji,2016. Query-focused multi-document summarization using hypergraph-based ranking. Information processing & management, vol. 52, no. 4, pp. 670–681, doi: 10.1016/J.IPM.2015.12.012.
45. [45] C. Republic, 2009. EVALUATION MEASURES FOR TEXT SUMMARIZATION Josef Steinberger , Karel Jezek. vol. 28, pp. 1001–1025.
46. [46] R. Mihalcea, 2005. Language independent extractive summarization. Proc. ACL 2005 Interact. poster Demonstr. Sess. - ACL ’05, no. June, pp. 49–52, doi: 10.3115/1225753.1225766.
47. [47] R. Mihalcea and P. Tarau, “TextRank: Bringing Order into Texts,” 1800.
48. [48] Lucy Vanderwende and Hisami Suzuki and Chris Brockett and Ani Nenkova2007. Beyond SumBasic: Task-focused summarization with sentence simplification and lexical expansion. Information processing & management, vol. 43, no. 6, pp. 1606–1618.
49. [49] A. Haghighi and L. Vanderwende,2009. Exploring content models for multi-document summarization. no. June, p. 362, doi: 10.3115/1620754.1620807.