Sound Natural: Content Rephrasing in Dialog Systems

TL;DR

This paper introduces a rephrasing task for dialog systems to enhance naturalness, presents a new dataset, and evaluates transformer and LSTM models, proposing a practical distilled LSTM solution.

Contribution

It defines a new rephrasing task for dialog systems, releases a dataset, and compares transformer and LSTM models, proposing an effective distilled LSTM approach.

Findings

BART with copy mechanisms performs strongly on rephrasing.

A distilled LSTM model offers a practical balance of performance and efficiency.

The dataset enables future research in natural dialog rephrasing.

Abstract

We introduce a new task of rephrasing for a more natural virtual assistant. Currently, virtual assistants work in the paradigm of intent slot tagging and the slot values are directly passed as-is to the execution engine. However, this setup fails in some scenarios such as messaging when the query given by the user needs to be changed before repeating it or sending it to another user. For example, for queries like 'ask my wife if she can pick up the kids' or 'remind me to take my pills', we need to rephrase the content to 'can you pick up the kids' and 'take your pills' In this paper, we study the problem of rephrasing with messaging as a use case and release a dataset of 3000 pairs of original query and rephrased query. We show that BART, a pre-trained transformers-based masked language model with auto-regressive decoding, is a strong baseline for the task, and show improvements by adding a copy-pointer and copy loss to it. We analyze different tradeoffs of BART-based and LSTM-based seq2seq models, and propose a distilled LSTM-based seq2seq as the best practical model.