End-to-End Conversational Search for Online Shopping with Utterance Transfer

TL;DR

This paper introduces ConvSearch, an end-to-end conversational search system for online shopping that uses utterance transfer to overcome data scarcity and outperforms baseline systems.

Contribution

It presents a novel ConvSearch system combining dialog and search, and proposes utterance transfer to generate training data from other domains and search logs.

Findings

Utterance transfer improves training data availability.

ConvSearch outperforms baseline systems.

Generated dataset enhances conversational shopping experience.

Abstract

Successful conversational search systems can present natural, adaptive and interactive shopping experience for online shopping customers. However, building such systems from scratch faces real word challenges from both imperfect product schema/knowledge and lack of training dialog data.In this work we first propose ConvSearch, an end-to-end conversational search system that deeply combines the dialog system with search. It leverages the text profile to retrieve products, which is more robust against imperfect product schema/knowledge compared with using product attributes alone. We then address the lack of data challenges by proposing an utterance transfer approach that generates dialogue utterances by using existing dialog from other domains, and leveraging the search behavior data from e-commerce retailer. With utterance transfer, we introduce a new conversational search dataset for online shopping. Experiments show that our utterance transfer method can significantly improve the availability of training dialogue data without crowd-sourcing, and the conversational search system significantly outperformed the best tested baseline.