Elamur Research⌕
☰
home / papers / 2015 / Cross-Document Co-Reference Resolution using Sample-Based Clustering with Knowledge Enrichment

Cross-Document Co-Reference Resolution using Sample-Based Clustering with Knowledge Enrichment

Summary

The paper tackles cross-document co-reference resolution (CCR): grouping entity mentions from many documents into equivalence classes, one per real-world entity. Unlike entity linking, CCR does not require entities to exist in a knowledge base. The authors propose CROCS, an unsupervised pipeline with four stages:

  1. It runs Stanford's sieve-based intra-document CR to form local mention chains.
  2. It enriches each chain with a "semantic summary" (semsum). The semsum combines context sentences with Freebase labels, drawn both for the chain itself and for co-occurring mentions of other entities, which helps disambiguate sparse names.
  3. It computes similarities lazily, only for sampled mention groups.
  4. It clusters through repeated top-down bisection. Each split uses sample-based spectral clustering or graph partitioning, and an information-theoretic criterion decides when splitting stops, so the number of entities need not be fixed in advance.

Experiments on three corpora report quality gains over the strongest earlier systems and much lower run-time. This lets CCR scale to large news collections where all-pairs similarity computation is impractical.

Key points

  • Evaluation covers the John Smith, WePS-2 and New York Times datasets, measured with B³ and φ3 metrics.
  • On WePS-2, full spectral clustering inside CROCS reaches B³ F1 ≈ 86.9 in about 331 s. The sampled spectral variant reaches ≈ 83.9 in about 86 s.
  • On John Smith, spectral clustering gives B³ F1 ≈ 79.9.
  • On New York Times, plain k-means had covered only about 5% of mentions when stopped at 20 hours. Sampled spectral clustering reaches F1 ≈ 62.2 in about 19.6 hours.
  • The preprocessing toolchain is Stanford CoreNLP, the Stanford NER Tagger and the Stanford multi-pass sieve CR. Freebase serves as the knowledge source.
  • The authors name multiple KBs and streaming input (news, social media) as future directions.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

Fields: Linguistics & languages

Topics: Coreference resolution

Cited by

8 papers of our library cite it (of the 140673 whose bibliographies we read); 1 of them have a page here, 7 not yet.

Cite

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.1162/tacl_a_00119

@article{dutta2015cross,
  title = {{Cross-Document Co-Reference Resolution using Sample-Based Clustering with Knowledge Enrichment}},
  author = {Dutta, Sourav and Weikum, Gerhard},
  journal = {Transactions of the Association for Computational Linguistics},
  year = {2015},
  doi = {10.1162/tacl\_a\_00119},
  url = {https://doi.org/10.1162/tacl_a_00119}
}

References

52 references, 17 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. J. Artiles, J. Gonzalo, S. Sekine: 2009. WePS 2 Evaluation Campaign: Overview of the Web People Search Clustering Task. find
  2. A. Bagga, B. Baldwin: 1998. Entity-Based Cross-Document Coreferencing Using the Vector Space Model. In COLING-ACL, pages 79–85. find
  3. C. F. Baker: 2012. FrameNet, Current Collaborations & Future Goals. LREC, 46(2):269–286. find
  4. A. Baron, M. Freedman: 2008. Who is Who & What is What: Experiments in Cross-Document Co-Reference. In EMNLP, pages 274–283. find
  5. A. Buluc, H. Meyerhenke, I. Safro, P. Sanders, C. Schulz: 2013. Recent Advances in Graph Partitioning. Karlsruhe Institute of Technology, Technical Report. find
  6. J. Cai, M. Strube: 2010. Evaluation Metrics for End-to-End Coreference Resolution Systems. In SIGDIAL, pages 28–36. find our card
  7. Y. Chen, J. Martin: 2007. Towards Robust Unsupervised Personal Name Disambiguation. In EMNLP-CoNLL, pages 190–198. find
  8. M. Cornolti, P. Ferragina, M. Ciaramita: 2013. A Framework for Benchmarking Entity-Annotation Systems. In WWW, pages 249–260. find
  9. S. Cucerzan: 2007. Large-Scale Named Entity Disambiguation Based on Wikipedia Data. In EMNLP-CoNLL, pages 708–716. find
  10. A. Culotta, M. L. Wick, A. McCallum: 2007. First-Order Probabilistic Models for Coreference Resolution. In HLT-NAACL, pages 81–88. find our card
  11. P. Domingos, S. Kok, D. Lowd, H. Poon, M. Richardson, P. Singla: 2007. Markov Logic. Probabilistic ILP. Springer-Verlag, pages 92–117. find
  12. P. Domingos, D. Lowd: 2009. Markov Logic: An Interface Layer for Artificial Intelligence. find
  13. Morgan and Claypool Publishers, 2009. find
  14. J. R. Finkel, T. Grenager, C. D. Manning: 2005. Incorporating Non-local Information into Information Extraction Systems by Gibbs Sampling. In ACL, pages 363–370. find
  15. C. H. Gooi, J. Allan: 2004. Cross-Document Coreference on a Large Scale Corpus. In HLT-NAACL, pages 9–16. find our card
  16. P. D. Grünwald: 2007. The Minimum Description Length Principle. MIT University Press. find
  17. A. Haghighi, D. Klein: 2009. Simple Coreference Resolution with Rich Syntactic and Semantic Features. In EMNLP, pages 1152–1161. find our card
  18. A. Haghighi, D. Klein: 2010. Coreference Resolution in a Modular, Entity-Centered Model. In HLT-NAACL, pages 385–393. find our card
  19. H. Hajishirzi, L. Zilles, D. S. Weld, L. S. Zettlemoyer: 2013. Joint Coreference Resolution and Named-Entity Linking with Multi-Pass Sieves. In EMNLP, pages 289–299. find our card
  20. J. Hoffart, M. A. Yosef, I. Bordino, H. Fürstenau, M. Pinkal, M. Spaniol, B. Taneva, S. Thater, G. Weikum: 2011. Robust Disambiguation of Named Entities in Text. In EMNLP, pages 782–792. find
  21. N. Hourdakis, M. Argyriou, G. M. Petrakis, E. E. Milios: 2010. Hierarchical Clustering in Medical Document Collections: the BIC-Means Method. Journal of Digital Information Management, 8(2):71–77. find
  22. G. Karypis, V. Kumar: 1998. A Fast and Highly Quality Multilevel Scheme for Partitioning Irregular Graphs. Journal on Scientific Computing, 20(1):359–392. find
  23. B. W. Kernighan, S. Lin: 1970. An efficient heuristic procedure for partitioning graphs. Bell System Technical Journal. find
  24. D. Koller, N. Friedman: 2009. Probabilistic Graphical Models: Principles and Techniques. MIT Press. find
  25. A. Krishnamurty, S. Balakrishnan, M. Xu, A. Singh: 2012. Efficient Active Algorithms for Hierarchical Clustering. In ICML, pages 887–894. find
  26. H. Lee, Y. Peirsman, A. Chang, N. Chambers, M. Surdeanu, D. Jurafsky: 2011. Stanford’s Multi-Pass Sieve Coreference Resolution System at the CoNLL-2011 Shared Task. In CoNLL, pages 28–34. find our card
  27. H. Lee, A. Chang, Y. Peirsman, N. Chambers, M. Surdeanu, D. Jurafsky: 2013. Deterministic Coreference Resolution based on Entity-centric, Precision-ranked Rules. Computational Linguistics Journal, 39(4):885– 916. find our card
  28. H. Lee, M. Recasens, A. X. Chang, M. Surdeanu, D. Jurafsky: 2012. Joint Entity and Event Coreference Resolution across Documents. In EMNLP-CoNLL, pages 489–500. find our card
  29. H. A. Loeliger: 2008. An Introduction to Factor Graphs. In MLSB. find
  30. X. Luo: 2005. On Coreference Resolution Performance Metrics. In HLT-EMNLP, pages 25–32. find our card
  31. U. von Luxburg: 2007. A Tutorial on Spectral Clustering. Statistics and Computing Journal, 17(4):395–416. find
  32. G. S. Mann, D. Yarowsky: 2003. Unsupervised Personal Name Disambiguation. In CoNLL,HLT-NAACL, pages 33–40. find
  33. D. N. Milne, I. H. Witten: 2008. Learning to Link with Wikipedia. In CIKM, pages 509–518. find
  34. V. Ng: 2010. Supervised Noun Phrase Coreference Research: The First Fifteen Years. In ACL, pages 1396– 1411. find our card
  35. C. Niu, W. Li, R. K. Srihari: 2004. Weakly Supervised Learning for Cross-document Person Name Disambiguation Supported by Information Extraction. In ACL, article 597. find
  36. K. Raghunathan, H. Lee, S. Rangarajan, N. Chambers, M. Surdeanu, D. Jurafsky, C. Manning: 2010. A Multi-Pass Sieve for Coreference Resolution. In EMNLP, pages 492–501. find our card
  37. A. Rahman, V. Ng: 2011a. Coreference Resolution with World Knowledge. In ACL, pages 814–824. find our card
  38. A. Rahman, V. Ng: 2011b. Ensemble-Based Coreference Resolution. In IJCAI, pages 1884–1889. find
  39. D. Rao, P. McNamee, M. Dredze: 2010. Streaming Cross Document Entity Coreference Resolution. In COL-ING, pages 1050–1058. find our card
  40. L. A. Ratinov, D. Roth, D. Downey, M. Anderson: 2011. Local and Global Algorithms for Disambiguation to Wikipedia. In ACL, pages 1375–1384 . find
  41. L. A. Ratinov, D. Roth: 2012. Learning-based Multi-Sieve Co-reference Resolution with Knowledge. In EMNLP-CoNLL, pages 1234–1244. find our card
  42. M. Richardson, P. Domingos: 2006. Markov Logic Networks. Journal of Machine Learning, 62(1-2):107– 136. find
  43. E. Sandhaus: 2008. The New York Times Annotated Corpus Overview. Linguistic Data Consortium. find
  44. G. E. Schwarz: 1978. Estimating the Dimension of a Model. Annals of Statistics, 6(2):461–464. find
  45. O. Shamir, N. Tishby: 2011. Spectral Clustering on a Budget. Journal of Machine Learning Research - Proceedings Track 15:661–669. find
  46. S. Singh, M. L. Wick, A. McCallum: 2010. Distantly Labeling Data for Large Scale Cross-Document Coreference. CoRR abs/1005.4298. — [arXiv:1005.4298](https://arxiv.org/abs/1005.4298) find
  47. S. Singh, A. Subramanya, F. Pereira, A. McCallum: 2011. Large-Scale Cross-Document Coreference Using Distributed Inference and Hierarchical Models. In ACL, pages 793–803. find our card
  48. F. M. Suchanek, G. Kasneci, G. Weikum: 2007. YAGO: a Core of Semantic Knowledge. In WWW, pages 697– 706. find
  49. B. Taneva, M. Kacimi, G. Weikum: 2011. Finding Images of Difficult Entities in the Long Tail. In CIKM, pages 189–194. find
  50. F. L. Wauthier, N. Jojic, M. I. Jordan: 2012. Active Spectral Clustering via Iterative Uncertainty Reduction. In KDD, pages 1339–1347. find
  51. D. Yan, L. Huang, M. I. Jordan: 2009. Fast approximate spectral clustering. In KDD, pages 907–916. find
  52. J. Zheng, L. Vilnis, S. Singh, J. D. Choi, A. McCallum: 2013. Dynamic Knowledge-Base Alignment for Coreference Resolution. In CoNLL, pages 153–162. find our card

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/papers/dutta-2015-cross-document-co-reference-resolution/