The assignment problem revisited

Alfaro, Carlos A.; Pérez, Sergio Luis Pérez; Valencia, Carlos E.; Vargas, Marcos C.

doi:10.1007/s11590-021-01791-4

Cited by 6 publications

(4 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the other mentions, a correct prediction based on Wikipedia entities is impossible. Instead, NIL-aware approaches could either (1) create an (intermediate) entity representation for the NIL entity to link, or (2) produce clusters of NIL mentions with all mentions in a cluster referring to the same entity.…”

Section: Motivation and Problemmentioning

confidence: 99%

“…With NASTyLinker, we present an EL approach that is NIL-aware in the sense of category (2) and hence avoiding the need for an adaptation dataset. Similar to Agarwal et al [1], it produces clusters of mentions and entities on the basis of inter-mention and mention-entity affinities from a bi-encoder.…”

Section: Approach and Contributionsmentioning

confidence: 99%

“…Instead, we only allow one-to-one mappings between cluster identifiers and NIL entities. Finding an optimal assignment for this scenario is equivalent to solving the linear sum assignment problem [2], for which efficient algorithms exist.…”

Section: Metricsmentioning

confidence: 99%

See 2 more Smart Citations

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

Section: Motivation and Problemmentioning

confidence: 99%

Section: Approach and Contributionsmentioning

confidence: 99%

See 1 more Smart Citation

NASTyLinker: NIL-Aware Scalable Transformer-based Entity Linker

Heist¹,

Paulheim²

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…To connect (5.14) to the assignment problem formulation [166] adopted in this chapter, we introduce its basic notation. Let L ∈ R C×K be a cost matrix whose entries L j,i represent the cost of assigning the j-th client to the i-th cluster, and A be a binary assignment matrix with entries A j,i ∈ {0, 1}.…”

Section: Data Association As An Optimization Problemmentioning

confidence: 99%