End-to-End Conversational Search for Online Shopping with Utterance Transfer
This work addresses the problem of data scarcity and imperfect product knowledge for developers building conversational search systems in e-commerce, though it is incremental as it builds on existing dialog and search techniques.
The authors tackled the challenge of building conversational search systems for online shopping by proposing ConvSearch, an end-to-end system that uses text profiles for robust product retrieval, and introduced an utterance transfer method to generate training data from other domains and e-commerce behavior, resulting in a new dataset and significant performance improvements over baselines.
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.