Structured Context and High-Coverage Grammar for Conversational Question Answering over Knowledge Graphs

TL;DR

This paper presents a neural semantic parsing approach with a new logical form grammar and structured Transformer input for conversational question answering over knowledge graphs, improving coverage and accuracy.

Contribution

Introduction of a new logical form grammar and a Transformer-based model that incorporates structured context for improved conversational QA over knowledge graphs.

Findings

Coverage increased from 80% to 96.2% on CSQA.

LF execution accuracy improved from 70.6% to 75.6%.

Achieved competitive results on ConvQuestions.

Abstract

We tackle the problem of weakly-supervised conversational Question Answering over large Knowledge Graphs using a neural semantic parsing approach. We introduce a new Logical Form (LF) grammar that can model a wide range of queries on the graph while remaining sufficiently simple to generate supervision data efficiently. Our Transformer-based model takes a JSON-like structure as input, allowing us to easily incorporate both Knowledge Graph and conversational contexts. This structured input is transformed to lists of embeddings and then fed to standard attention layers. We validate our approach, both in terms of grammar coverage and LF execution accuracy, on two publicly available datasets, CSQA and ConvQuestions, both grounded in Wikidata. On CSQA, our approach increases the coverage from $80\%$ to $96.2\%$, and the LF execution accuracy from $70.6\%$ to $75.6\%$, with respect to previous state-of-the-art results. On ConvQuestions, we achieve competitive results with respect to the state-of-the-art.