We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs

Change to browse by:

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Information Retrieval

Title: Entity-aware Transformers for Entity Search

Abstract: Pre-trained language models such as BERT have been a key ingredient to achieve state-of-the-art results on a variety of tasks in natural language processing and, more recently, also in information retrieval.Recent research even claims that BERT is able to capture factual knowledge about entity relations and properties, the information that is commonly obtained from knowledge graphs. This paper investigates the following question: Do BERT-based entity retrieval models benefit from additional entity information stored in knowledge graphs? To address this research question, we map entity embeddings into the same input space as a pre-trained BERT model and inject these entity embeddings into the BERT model. This entity-enriched language model is then employed on the entity retrieval task. We show that the entity-enriched BERT model improves effectiveness on entity-oriented queries over a regular BERT model, establishing a new state-of-the-art result for the entity retrieval task, with substantial improvements for complex natural language queries and queries requesting a list of entities with a certain property. Additionally, we show that the entity information provided by our entity-enriched model particularly helps queries related to less popular entities. Last, we observe empirically that the entity-enriched BERT models enable fine-tuning on limited training data, which otherwise would not be feasible due to the known instabilities of BERT in few-sample fine-tuning, thereby contributing to data-efficient training of BERT for entity search.
Subjects: Information Retrieval (cs.IR); Computation and Language (cs.CL)
ACM classes: H.3.3
DOI: 10.1145/3477495.3531971
Cite as: arXiv:2205.00820 [cs.IR]
  (or arXiv:2205.00820v1 [cs.IR] for this version)

Submission history

From: Faegheh Hasibi [view email]
[v1] Mon, 2 May 2022 11:53:59 GMT (585kb,D)

Link back to: arXiv, form interface, contact.