Learning to Ask: Conversational Product Search via Representation Learning

TL;DR

The proposed ConvPS model can naturally integrate the representation learning of the user, query, item, and conversation into a unified generative framework, which provides a promising avenue for constructing accurate and robust conversational product search systems that are flexible and adaptive.

Abstract

Online shopping platforms, such as Amazon and AliExpress, are increasingly prevalent in society, helping customers purchase products conveniently. With recent progress in natural language processing, researchers and practitioners shift their focus from traditional product search to conversational product search. Conversational product search enables user-machine conversations and through them collects explicit user feedback that allows to actively clarify the users' product preferences. Therefore, prospective research on an intelligent shopping assistant via conversations is indispensable. Existing publications on conversational product search either model conversations independently from users, queries, and products or lead to a vocabulary mismatch. In this work, we propose a new conversational product search model, ConvPS, to assist users in locating desirable items. The model is first trained to jointly learn the semantic representations of user, query, item, and conversation via a unified generative framework. After learning these representations, they are integrated to retrieve the target items in the latent semantic space. Meanwhile, we propose a set of greedy and explore-exploit strategies to learn to ask the user a sequence of high-performance questions for conversations. Our proposed ConvPS model can naturally integrate the representation learning of the user, query, item, and conversation into a unified generative framework, which provides a promising avenue for constructing accurate and robust conversational product search systems that are flexible and adaptive. Experimental results demonstrate that our ConvPS model significantly outperforms state-of-the-art baselines.

Figures (6)

• Figure 1: The research framework of our ConvPS model. (1) We first construct a question pool via slot-value pair extraction. (2) We then learn the user embeddings, conversation embeddings in terms of slot-value embeddings, item embeddings, and word embeddings to formulate query embeddings via our generative model. (3) Afterwards, we learn to ask a sequence of questions by the four proposed strategies: GBS strategy, LinRel strategy, GP+EI strategy, and GP+UCB strategy. (4) Lastly, we generate the item ranked list by estimating the probability of items based on these learned embeddings.
• Figure 2: Performance of different question selection strategies.
• Figure 3: The performance of ConvPS with different components removed.
• Figure 4: Effect of embedding sizes.
• Figure 5: Effect of batch sizes.