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.

References in corpus (11)

• Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
• MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
• Autoregressive Entity Retrieval
• Multi-Stage Document Ranking with BERT
• Word-Entity Duet Representations for Document Ranking
• Investigating Entity Knowledge in BERT with Simple Neural End-To-End Entity Linking
• Language Models are Open Knowledge Graphs
• Inductive Entity Representations from Text via Link Prediction
• Improving Efficient Neural Ranking Models with Cross-Architecture Knowledge Distillation
• Dagstuhl Perspectives Workshop 24352 -- Conversational Agents: A Framework for Evaluation (CAFE): Manifesto
• Graph-Embedding Empowered Entity Retrieval

Cited by in corpus (3)

• Fine Tuning vs. Retrieval Augmented Generation for Less Popular Knowledge
• Personal Entity, Concept, and Named Entity Linking in Conversations
• MMEAD: MS MARCO Entity Annotations and Disambiguations