Entity-aware Transformers for Entity Search

TL;DR

This paper introduces an entity-enriched BERT model that incorporates knowledge graph information, significantly improving entity retrieval performance, especially for complex, less popular, and data-scarce queries.

Contribution

The paper proposes a novel method to inject knowledge graph-based entity embeddings into BERT, enhancing its effectiveness and data efficiency for entity search tasks.

Findings

Improves entity retrieval effectiveness over standard BERT.

Achieves state-of-the-art results on entity retrieval benchmarks.

Enables effective fine-tuning with limited training data.

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.