An End-to-End Re-Evaluation of Table Entity-Linkers

Incomplete entity linker evaluation: Existing evaluations of table-to-KG entity linkers are incomplete because they focus on specific applications and aggregate output metrics rather than component effectiveness and scalability.

Entity linker reproducibility: Many entity linkers are difficult to reproduce due to unavailable source code or dependence on irreproducible public endpoints and datasets.

Incomplete entity linker evaluation: Existing evaluations of table-to-KG entity linkers are incomplete because they focus on specific applications and aggregate output metrics rather than component effectiveness and scalability.

Entity linker reproducibility: Many entity linkers are difficult to reproduce due to unavailable source code or dependence on irreproducible public endpoints and datasets.

End-to-end component evaluation: The paper proposes an in-depth evaluation of state-of-the-art entity linkers that assesses the quality and scalability of individual entity linking components.

Multi-benchmark KG linker evaluation: The paper evaluates entity linkers on four existing entity linking benchmarks using DBpedia and Wikidata.

End-to-end component evaluation: The paper proposes an in-depth evaluation of state-of-the-art entity linkers that assesses the quality and scalability of individual entity linking components.

Multi-benchmark KG linker evaluation: The paper evaluates entity linkers on four existing entity linking benchmarks using DBpedia and Wikidata.

Bottleneck identification:

Candidate generation is crucial:

Entity linkers often irreproducible: