Answer Candidate Type Selection: Text-to-Text Language Model for Closed Book Question Answering Meets Knowledge Graphs

TL;DR

This paper introduces a method to improve knowledge graph question answering by filtering and re-ranking generated answers based on entity types from Wikidata, enhancing accuracy especially for less popular entities.

Contribution

It proposes a novel type-based filtering and re-ranking approach that enhances pre-trained text-to-text models for KGQA tasks.

Findings

Improved answer accuracy for less popular entities

Effective filtering and re-ranking based on entity types

Enhancement over baseline pre-trained models

Abstract

Pre-trained Text-to-Text Language Models (LMs), such as T5 or BART yield promising results in the Knowledge Graph Question Answering (KGQA) task. However, the capacity of the models is limited and the quality decreases for questions with less popular entities. In this paper, we present a novel approach which works on top of the pre-trained Text-to-Text QA system to address this issue. Our simple yet effective method performs filtering and re-ranking of generated candidates based on their types derived from Wikidata "instance_of" property.