Providing relevant, diverse, and fair results is crucial for informationretrieval systems. It has attracted more and more attentionbecause of issues caused by traditional relevance-centric retrieval systems.These issues include the problem of echo chambers and theincreasingly polarized online communities. Therefore, we participatedin the NTCIR-17 FairWeb-1 Task to provide group fairness toresearchers, movies, and YouTube content and submitted five runs.The runs are based on a recently proposed fair ranking framework,DLF. The experimental results demonstrate that, in many cases, DLF can improve fairness while maintaining relevance but stillneeds more exploration for ordinal fairness groups and documentswith longer text. This paper reports how the runs were constructedand discusses their performance and future work.
Chen et al. (2023) studied this question.