Visual Question Rewriting for Increasing Response Rate

TL;DR

This paper introduces Visual Question Rewriting (VQR), a new task that uses visual information to automatically rewrite questions to be more attractive and increase response rates, supported by a new dataset and baseline models.

Contribution

The paper presents the VQR task, a new dataset of questions with images, and baseline models demonstrating the effectiveness of visual cues in question rewriting.

Findings

Rewritten questions are more attractive and elicit higher response rates.

Images contribute positively to the quality of rewritten questions.

Baseline models can effectively incorporate visual information for question rewriting.

Abstract

When a human asks questions online, or when a conversational virtual agent asks human questions, questions triggering emotions or with details might more likely to get responses or answers. we explore how to automatically rewrite natural language questions to improve the response rate from people. In particular, a new task of Visual Question Rewriting(VQR) task is introduced to explore how visual information can be used to improve the new questions. A data set containing around 4K bland questions, attractive questions and images triples is collected. We developed some baseline sequence to sequence models and more advanced transformer based models, which take a bland question and a related image as input and output a rewritten question that is expected to be more attractive. Offline experiments and mechanical Turk based evaluations show that it is possible to rewrite bland questions in a more detailed and attractive way to increase the response rate, and images can be helpful.