A collective approach to ranking entities for mentions

Shunlin Rong, Mizuho Iwaihara

Research output: Chapter in Book/Report/Conference proceedingConference contribution

1 Citation (Scopus)

Abstract

Entity linking (EL) is the task of mapping name mentions in web text to their entities in a knowledge base. Most of earlier EL work in the knowledge based approach is usually formulated as a ranking problem, either by (i) non-collective approaches with supervised models, or (ii) collective approaches by leveraging global topical coherence which means semantic relations between entities through graph-based approaches. For the mapping process, we can regard it as selecting an entity to its mention by combining these two methods. In this paper, we propose a probabilistic model that ranks related entities to name mentions where ranking is customized by using three types of data: popularity knowledge of the entity, context similarity between mentions and the entity, and semantic relations between mapping entities. Specifically, we first propose an EL model utilizing global topical coherence that means semantic relatedness between entities, as well as using local mention-to-entity compatibility, to improve recall and precision. The key benefit of our model comes from 1) combination of two methods to provide customized ranking for mentions, 2) the model can save a large amount of calculation by efficiently finding candidate combinations of entities through global semantic coherence.

Original languageEnglish
Title of host publication2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509008063
DOIs
Publication statusPublished - 2016 Aug 23
Event15th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2016 - Okayama, Japan
Duration: 2016 Jun 262016 Jun 29

Other

Other15th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2016
CountryJapan
CityOkayama
Period16/6/2616/6/29

Fingerprint

Ranking
Semantics
Linking
Knowledge-based
Probabilistic Model
Knowledge Base
Model
Compatibility
Graph in graph theory

Keywords

  • Entity Linking
  • Ranking
  • Wikipedia

ASJC Scopus subject areas

  • Computer Science(all)
  • Energy Engineering and Power Technology
  • Control and Optimization

Cite this

Rong, S., & Iwaihara, M. (2016). A collective approach to ranking entities for mentions. In 2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings [7550859] Institute of Electrical and Electronics Engineers Inc.. https://doi.org/10.1109/ICIS.2016.7550859

A collective approach to ranking entities for mentions. / Rong, Shunlin; Iwaihara, Mizuho.

2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2016. 7550859.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Rong, S & Iwaihara, M 2016, A collective approach to ranking entities for mentions. in 2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings., 7550859, Institute of Electrical and Electronics Engineers Inc., 15th IEEE/ACIS International Conference on Computer and Information Science, ICIS 2016, Okayama, Japan, 16/6/26. https://doi.org/10.1109/ICIS.2016.7550859
Rong S, Iwaihara M. A collective approach to ranking entities for mentions. In 2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc. 2016. 7550859 https://doi.org/10.1109/ICIS.2016.7550859
Rong, Shunlin ; Iwaihara, Mizuho. / A collective approach to ranking entities for mentions. 2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings. Institute of Electrical and Electronics Engineers Inc., 2016.
@inproceedings{d7f0713290814159a0ca8dbaacc98185,
title = "A collective approach to ranking entities for mentions",
abstract = "Entity linking (EL) is the task of mapping name mentions in web text to their entities in a knowledge base. Most of earlier EL work in the knowledge based approach is usually formulated as a ranking problem, either by (i) non-collective approaches with supervised models, or (ii) collective approaches by leveraging global topical coherence which means semantic relations between entities through graph-based approaches. For the mapping process, we can regard it as selecting an entity to its mention by combining these two methods. In this paper, we propose a probabilistic model that ranks related entities to name mentions where ranking is customized by using three types of data: popularity knowledge of the entity, context similarity between mentions and the entity, and semantic relations between mapping entities. Specifically, we first propose an EL model utilizing global topical coherence that means semantic relatedness between entities, as well as using local mention-to-entity compatibility, to improve recall and precision. The key benefit of our model comes from 1) combination of two methods to provide customized ranking for mentions, 2) the model can save a large amount of calculation by efficiently finding candidate combinations of entities through global semantic coherence.",
keywords = "Entity Linking, Ranking, Wikipedia",
author = "Shunlin Rong and Mizuho Iwaihara",
year = "2016",
month = "8",
day = "23",
doi = "10.1109/ICIS.2016.7550859",
language = "English",
booktitle = "2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
address = "United States",

}

TY - GEN

T1 - A collective approach to ranking entities for mentions

AU - Rong, Shunlin

AU - Iwaihara, Mizuho

PY - 2016/8/23

Y1 - 2016/8/23

N2 - Entity linking (EL) is the task of mapping name mentions in web text to their entities in a knowledge base. Most of earlier EL work in the knowledge based approach is usually formulated as a ranking problem, either by (i) non-collective approaches with supervised models, or (ii) collective approaches by leveraging global topical coherence which means semantic relations between entities through graph-based approaches. For the mapping process, we can regard it as selecting an entity to its mention by combining these two methods. In this paper, we propose a probabilistic model that ranks related entities to name mentions where ranking is customized by using three types of data: popularity knowledge of the entity, context similarity between mentions and the entity, and semantic relations between mapping entities. Specifically, we first propose an EL model utilizing global topical coherence that means semantic relatedness between entities, as well as using local mention-to-entity compatibility, to improve recall and precision. The key benefit of our model comes from 1) combination of two methods to provide customized ranking for mentions, 2) the model can save a large amount of calculation by efficiently finding candidate combinations of entities through global semantic coherence.

AB - Entity linking (EL) is the task of mapping name mentions in web text to their entities in a knowledge base. Most of earlier EL work in the knowledge based approach is usually formulated as a ranking problem, either by (i) non-collective approaches with supervised models, or (ii) collective approaches by leveraging global topical coherence which means semantic relations between entities through graph-based approaches. For the mapping process, we can regard it as selecting an entity to its mention by combining these two methods. In this paper, we propose a probabilistic model that ranks related entities to name mentions where ranking is customized by using three types of data: popularity knowledge of the entity, context similarity between mentions and the entity, and semantic relations between mapping entities. Specifically, we first propose an EL model utilizing global topical coherence that means semantic relatedness between entities, as well as using local mention-to-entity compatibility, to improve recall and precision. The key benefit of our model comes from 1) combination of two methods to provide customized ranking for mentions, 2) the model can save a large amount of calculation by efficiently finding candidate combinations of entities through global semantic coherence.

KW - Entity Linking

KW - Ranking

KW - Wikipedia

UR - http://www.scopus.com/inward/record.url?scp=84988009559&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988009559&partnerID=8YFLogxK

U2 - 10.1109/ICIS.2016.7550859

DO - 10.1109/ICIS.2016.7550859

M3 - Conference contribution

BT - 2016 IEEE/ACIS 15th International Conference on Computer and Information Science, ICIS 2016 - Proceedings

PB - Institute of Electrical and Electronics Engineers Inc.

ER -