Does Structure Matter? Leveraging Data-to-Text Generation for Answering Complex Information Needs

TL;DR

This paper explores how structured data-to-text generation can improve answers to complex questions by planning and content selection, demonstrating its effectiveness over traditional text-to-text models on the TREC CAR dataset.

Contribution

It introduces a content planning pipeline for data-to-text generation tailored for complex answer generation, showing improvements over standard models.

Findings

Planning-based models outperform text-to-text models

Structured answers improve relevance and coherence

Effective content selection enhances answer quality

Abstract

In this work, our aim is to provide a structured answer in natural language to a complex information need. Particularly, we envision using generative models from the perspective of data-to-text generation. We propose the use of a content selection and planning pipeline which aims at structuring the answer by generating intermediate plans. The experimental evaluation is performed using the TREC Complex Answer Retrieval (CAR) dataset. We evaluate both the generated answer and its corresponding structure and show the effectiveness of planning-based models in comparison to a text-to-text model.