Learning Latent Vector Spaces for Product Search

TL;DR

This paper presents a new latent vector space model that jointly learns representations of words and products in e-commerce, improving product search by capturing their discriminative relations without explicit annotations.

Contribution

The paper introduces a novel joint learning approach for word and product embeddings that enhances product search performance without relying on labeled data.

Findings

Outperforms existing latent vector models like LSI, LDA, and word2vec.

Learns better product representations through joint modeling.

Mapping from words to products benefits from backpropagated errors.

Abstract

We introduce a novel latent vector space model that jointly learns the latent representations of words, e-commerce products and a mapping between the two without the need for explicit annotations. The power of the model lies in its ability to directly model the discriminative relation between products and a particular word. We compare our method to existing latent vector space models (LSI, LDA and word2vec) and evaluate it as a feature in a learning to rank setting. Our latent vector space model achieves its enhanced performance as it learns better product representations. Furthermore, the mapping from words to products and the representations of words benefit directly from the errors propagated back from the product representations during parameter estimation. We provide an in-depth analysis of the performance of our model and analyze the structure of the learned representations.