An Attention-Based Interactive Learning-to-Rank Model for Document Retrieval
Fan Zhang,Wenyu Chen,Mingsheng Fu,Fan Li,Hong Qu,Zhang Yi
DOI: https://doi.org/10.1109/TSMC.2021.3129839
2022-01-01
IEEE Transactions on Systems, Man, and Cybernetics: Systems
Abstract:The core issue of learning-to-rank (LTR) for document retrieval lies in finding an optimal ranking policy to meet the search intent of the user. The majority of proposed LTR approaches treat the ranking as a static process, employing a fixed ranking policy to immediately assign scores to documents. By contrast, ranking is not a static but an interactive process where the user continues interacting with the document retrieval system through information exchange such as search intent (e.g., rating or clicking for the retrieved items). We model the interactive ranking process (IRP), and propose an Attention-Based Interactive LTR model (AIRank) to constitute an intent-aware flexible ranking policy to gratify the user's need. To enhance the ranking quality, the inherent relations among documents are procured by the self-attention method to contribute to an enriched user intent representation. Furthermore, we mend the policy gradient learning method to train the AIRank in the IRP. Experiments demonstrate the effectiveness of AIRank compared to the state-of-the-art methods in terms of normalized discounted cumulative gain and expected reciprocal rank.