Benchmarking Scalable Methods for Streaming Cross Document Entity Coreference

Logan, Robert L.; McCallum, Andrew; Singh, Sameer; Bikel, Dan

doi:10.18653/v1/2021.acl-long.364

Cited by 7 publications

(14 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…NIL clustering aims at grouping together mentions referring to the same entity. Several algorithms have been proposed, based on GNN ( [8]) or hierarchical clustering ( [9]). Incremental management to add NIL mentions to a KB has been considered in [10]…”

Section: Nil Mentions Managementmentioning

confidence: 99%

An Entity Registry: A Model for a Repository of Entities Found in a Document Set

Bellandi¹,

Siccardi²

2023

Natural Language Processing, Information Retrieval and AI

View full text Add to dashboard Cite

This paper proposes a conceptual structure for a repository of entities that can be found by usual procedures of Natural Language Processing, that is the search for entities mentioned in text, their identification, possibly through the link to entries in Background Knowledge Basis (BKG) and theconstruction of a Knowledge Basis or Graph to host the information found in this process. We address applications where a BKG is of little help, because the involved entities are not so relevant to be included in any, being for instance ordinary people or small companies. Therefore, we rely on the entities’ attributes and relationships for unique identification, disambiguation, knowledge checking and any other relevant operation. One of the final goals achieved by the proposed method is the ability to merge knowledge collected in separate bases, once they are referred to the same Entity Registry.

show abstract

Section: Nil Mentions Managementmentioning

confidence: 99%

An Entity Registry: A Model for a Repository of Entities Found in a Document Set

Bellandi¹,

Siccardi²

2023

Natural Language Processing, Information Retrieval and AI

View full text Add to dashboard Cite

show abstract

“…Dutta and Weikum [11] explicitly tackle CDC in combination with EL by applying clustering to bag-of-words representations of entity mentions. More recently, Logan IV et al [25] evaluate greedy nearest-neighbour and hierarchical clustering strategies for CDC, however, without explicitly evaluating them with respect to EL.…”

Section: Related Workmentioning

confidence: 99%

“…To produce an initial mention clustering, we follow Logan IV et al [25] and use a greedy nearest-neighbour clustering. Given the mention affinity threshold τ m , the mentions M are grouped into clusters C so that two mentions m, m ∈ M belong to the same cluster if φ(m, m ) > τ m .…”

Section: Cluster Initializationmentioning

confidence: 99%

NASTyLinker: NIL-Aware Scalable Transformer-based Entity Linker

Heist¹,

Paulheim²

2023

Preprint

View full text Add to dashboard Cite

Entity Linking (EL) is the task of detecting mentions of entities in text and disambiguating them to a reference knowledge base. Most prevalent EL approaches assume that the reference knowledge base is complete. In practice, however, it is necessary to deal with the case of linking to an entity that is not contained in the knowledge base (NIL entity). Recent works have shown that, instead of focusing only on affinities between mentions and entities, considering inter-mention affinities can be used to represent NIL entities by producing clusters of mentions. At the same time, inter-mention affinities can help to substantially improve linking performance for known entities. With NASTyLinker, we introduce an EL approach that is aware of NIL entities and produces corresponding mention clusters while maintaining high linking performance for known entities. The approach clusters mentions and entities based on dense representations from Transformers and resolves conflicts (if more than one entity is assigned to a cluster) by computing transitive mention-entity affinities. We show the effectiveness and scalability of NASTyLinker on NILK, a dataset that is explicitly constructed to evaluate EL with respect to NIL entities. Further, we apply the presented approach to an actual EL task, namely to knowledge graph population by linking entities in Wikipedia listings, and provide an analysis of the outcome.

show abstract

“…Most of the related work on cross-document IE has focused on coreference resolution task [277][278][279][280][281][282][283][284][285][286][287][288]. This task consists in identifying coreferent mentions on a set of documents given as input.…”

Section: Effect Of Pre-training On New Corporamentioning

confidence: 99%

Towards Consistent Document-level Entity Linking: Joint Models for Entity Linking and Coreference Resolution

Zaporojets

Deleu²,

Demeester³

et al. 2022

Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 2: Short Papers)

View full text Add to dashboard Cite

We consider the task of document-level entity linking (EL), where it is important to make consistent decisions for entity mentions over the full document jointly. We aim to leverage explicit "connections" among mentions within the document itself: we propose to join EL and coreference resolution (coref) in a single structured prediction task over directed trees and use a globally normalized model to solve it. This contrasts with related works where two separate models are trained for each of the tasks and additional logic is required to merge the outputs. Experimental results on two datasets show a boost of up to +5% F1score on both coref and EL tasks, compared to their standalone counterparts. For a subset of hard cases, with individual mentions lacking the correct EL in their candidate entity list, we obtain a +50% increase in accuracy. 1

show abstract

Benchmarking Scalable Methods for Streaming Cross Document Entity Coreference

Cited by 7 publications

References 19 publications

An Entity Registry: A Model for a Repository of Entities Found in a Document Set

An Entity Registry: A Model for a Repository of Entities Found in a Document Set

NASTyLinker: NIL-Aware Scalable Transformer-based Entity Linker

Towards Consistent Document-level Entity Linking: Joint Models for Entity Linking and Coreference Resolution

Contact Info

Product

Resources

About