Abstract
Entity disambiguation, or mapping a phrase to its canonical representation in a knowledge base, is a fundamental step in many natural language processing applications. Existing techniques based on global ranking models fail to capture the individual peculiarities of the words and hence, struggle to meet the accuracy-time requirements of many real-world applications. In this paper, we propose a new system that learns specialized features and models for disambiguating each ambiguous phrase in the English language. We train and validate the hundreds of thousands of learning models for this purpose using a Wikipedia hyperlink dataset with more than 170 million labelled annotations. The computationally intensive training required for this approach can be distributed over a cluster. In addition, our approach supports fast queries, efficient updates and its accuracy compares favorably with respect to other state-of-the-art disambiguation systems.
T. Mai—Now at TrustingSocial (tiep@trustingsocial.com).
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
We used WikiExtractor (http://medialab.di.unipi.it/wiki/Wikipedia_Extractor) on the 2015-07-29 dump.
- 2.
In our notation, a sense is a Wikipedia entity and is coupled with a specific mention.
- 3.
- 4.
- 5.
- 6.
We used Spotlight 0.7 [4] (statistical model en_2+2 with the SpotXmlParser.
- 7.
We used the TAGME version 1.8 web API http://tagme.di.unipi.it/tag in January, 2016.
- 8.
- 9.
See the main Gerbil website as well as https://github.com/AKSW/gerbil/wiki/D2KB#handling-of-higher-order-annotators for more details. To quote the GERBIL documentation, “The response of these annotators is filtered using a strong annotation match filter. Thus, all entities that do not exactly match one of the marked entities in the gold standard are removed from the response of the annotator before it is evaluated.”.
- 10.
Ideally, to achieve better performance, one would need to adapt and retrain supervised models for scenarios with short and dynamic contexts such as KORE50 dataset. One potential issue of such retraining is the lack of big labelled data. This issue could be solved by integrating the target labelled dataset with Wikipedia dataset and adjusting the sample weights to balance the training cost of the target and Wikipedia datasets. However, we decided not to do so to maintain the fairness of this comparison.
References
Brando, C., Frontini, F., Ganascia, J.: REDEN: named entity linking in digital literary editions using linked data sets. CSIMQ 7, 60–80 (2016)
Cucerzan, S.: Large-scale named entity disambiguation based on wikipedia data. In: Proceedings of the EMNLP-CoNLL, pp. 708–716, June 2007
Cucerzan, S.: Name entities made obvious: the participation in the ERD 2014 evaluation. In: Proceedings of the ERD, pp. 95–100. ACM, New York (2014)
Daiber, J., Jakob, M., Hokamp, C., Mendes, P.N.: Improving efficiency and accuracy in multilingual entity extraction. In: Proceedings of the I-SEMANTICS (2013)
Ferragina, P., Scaiella, U.: TAGME: on-the-fly annotation of short text fragments (by Wikipedia entities). In: Proceedings of the CIKM, pp. 1625–1628 (2010)
Ferrucci, D.A.: Introduction to “This is Watson”. IBM J. Res. Dev. 56(3), 235–249 (2012)
Ganea, O., Ganea, M., Lucchi, A., Eickhoff, C., Hofmann, T.: Probabilistic bag-of-hyperlinks model for entity linking. In: Proceedings of the WWW, pp. 927–938 (2016)
Guo, Z., Barbosa, D.: Robust entity linking via random walks. In: Proceedings of the CIKM, pp. 499–508 (2014)
Han, X., Sun, L.: A generative entity-mention model for linking entities with knowledge base. In: Proceedings of the HLT, pp. 945–954 (2011)
Han, X., Sun, L., Zhao, J.: Collective entity linking in web text: a graph-based method. In: Proceedings of the SIGIR, pp. 765–774 (2011)
Hoffart, J.: Discovering and disambiguating named entities in text. In: Proceedings of the SIGMOD/PODS Ph.D. Symposium, pp. 43–48 (2013)
Houlsby, N., Ciaramita, M.: A scalable Gibbs sampler for probabilistic entity linking. In: Rijke, M., Kenter, T., Vries, A.P., Zhai, C.X., Jong, F., Radinsky, K., Hofmann, K. (eds.) ECIR 2014. LNCS, vol. 8416, pp. 335–346. Springer, Cham (2014). doi:10.1007/978-3-319-06028-6_28
Hulpuş, I., Prangnawarat, N., Hayes, C.: Path-based semantic relatedness on linked data and its use to word and entity disambiguation. In: Arenas, M., Corcho, O., Simperl, E., Strohmaier, M., d’Aquin, M., Srinivas, K., Groth, P., Dumontier, M., Heflin, J., Thirunarayan, K., Staab, S. (eds.) ISWC 2015. LNCS, vol. 9366, pp. 442–457. Springer, Cham (2015). doi:10.1007/978-3-319-25007-6_26
Kulkarni, S., Singh, A., Ramakrishnan, G., Chakrabarti, S.: Collective annotation of Wikipedia entities in web text. In: Proceedings of the KDD, pp. 457–466 (2009)
McNamee, P.: HLTCOE efforts in entity linking at TAC KBP 2010. In: Proceedings of the TAC (2010)
Meij, E., Weerkamp, W., de Rijke, M.: Adding semantics to microblog posts. In: Proceedings of the WSDM, pp. 563–572 (2012)
Milne, D., Witten, I.H.: Learning to link with Wikipedia. In: Proceedings of the CIKM, pp. 509–518 (2008)
Moro, A., Raganato, A., Navigli, R.: Entity linking meets word sense disambiguation: a unified approach. TACL 2, 231–244 (2014)
Olieman, A., Azarbonyad, H., Dehghani, M., Kamps, J., Marx, M.: Entity linking by focusing DBpedia candidate entities. In: Proceedings of the ERD, pp. 13–24 (2014)
Piccinno, F., Ferragina, P.: From TAGME to WAT: a new entity annotator. In: Proceedings of the ERD, pp. 55–62 (2014)
Qureshi, M.A., O’Riordan, C., Pasi, G.: Exploiting wikipedia for entity name disambiguation in tweets. In: Proceedings of the NLDB, pp. 184–195 (2014)
Suchanek, F., Weikum, G.: Knowledge harvesting in the big-data era. In: Proceedings of the SIGMOD, pp. 933–938. ACM, New York
Usbeck, R., Ngomo, A.N., Röder, M., Gerber, D., Coelho, S.A., Auer, S., Both, A.: AGDISTIS - agnostic disambiguation of named entities using linked open data. In: Proceedings of the ECAI, pp. 1113–1114 (2014)
Usbeck, R., Röder, M., Ngonga Ngomo, A.-C., Baron, C., Both, A., Brümmer, M., Ceccarelli, D., Cornolti, M., Cherix, D., Eickmann, B., Ferragina, P., Lemke, C., Moro, A., Navigli, R., Piccinno, F., Rizzo, G., Sack, H., Speck, R., Troncy, R., Waitelonis, J., Wesemann, L.: GERBIL: general entity annotator benchmarking framework. In: Proceedings of the WWW, pp. 1133–1143 (2015)
Zwicklbauer, S., Seifert, C., Granitzer, M.: Robust and collective entity disambiguation through semantic embeddings. In: Proceedings of the 39th International ACM SIGIR Conference on Research and Development in Information Retrieval, pp. 425–434. ACM (2016)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2017 Springer International Publishing AG
About this paper
Cite this paper
Mai, T., Shi, B., Nicholson, P.K., Ajwani, D., Sala, A. (2017). Scalable Disambiguation System Capturing Individualities of Mentions. In: Gracia, J., Bond, F., McCrae, J., Buitelaar, P., Chiarcos, C., Hellmann, S. (eds) Language, Data, and Knowledge. LDK 2017. Lecture Notes in Computer Science(), vol 10318. Springer, Cham. https://doi.org/10.1007/978-3-319-59888-8_31
Download citation
DOI: https://doi.org/10.1007/978-3-319-59888-8_31
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-59887-1
Online ISBN: 978-3-319-59888-8
eBook Packages: Computer ScienceComputer Science (R0)