Personal Entity, Concept, and Named Entity Linking in Conversations

TL;DR

This paper introduces CREL, a novel entity linking toolkit designed for conversations, which outperforms existing methods by effectively linking named entities, concepts, and personal entities using a new dataset and coreference resolution.

Contribution

The paper presents a new conversational entity linking dataset and a toolkit, CREL, that improves linking accuracy by handling personal entities and concepts with coreference resolution.

Findings

CREL outperforms state-of-the-art EL methods in conversational settings.

A new dataset of 1327 annotated conversational utterances was created.

CREL effectively links personal entities, concepts, and named entities.

Abstract

Building conversational agents that can have natural and knowledge-grounded interactions with humans requires understanding user utterances. Entity Linking (EL) is an effective and widely used method for understanding natural language text and connecting it to external knowledge. It is, however, shown that existing EL methods developed for annotating documents are suboptimal for conversations, where personal entities (e.g., "my cars") and concepts are essential for understanding user utterances. In this paper, we introduce a collection and a tool for entity linking in conversations. We collect EL annotations for 1327 conversational utterances, consisting of links to named entities, concepts, and personal entities. The dataset is used for training our toolkit for conversational entity linking, CREL. Unlike existing EL methods, CREL is developed to identify both named entities and concepts. It also utilizes coreference resolution techniques to identify personal entities and references to the explicit entity mentions in the conversations. We compare CREL with state-of-the-art techniques and show that it outperforms all existing baselines.