Reference Product Search
This work addresses product search for downstream modeling and business applications, but it appears incremental as it builds on existing embedding and retrieval techniques.
The paper tackles the problem of searching for reference products by proposing a method that combines product representation learning and fingerprint-type vector searching, achieving improved search return rate and precision compared to peer services.
For a product of interest, we propose a search method to surface a set of reference products. The reference products can be used as candidates to support downstream modeling tasks and business applications. The search method consists of product representation learning and fingerprint-type vector searching. The product catalog information is transformed into a high-quality embedding of low dimensions via a novel attention auto-encoder neural network, and the embedding is further coupled with a binary encoding vector for fast retrieval. We conduct extensive experiments to evaluate the proposed method, and compare it with peer services to demonstrate its advantage in terms of search return rate and precision.