Ranking-based Fusion Algorithms for Extreme Multi-label Text Classification (XMTC)

In the context of Extreme Multi-label Text Classification (XMTC), where labels are assigned to text instances from a large label space, the long-tail distribution of labels presents a significant challenge. Labels can be broadly categorized into frequent, high-coverage $\mathbf{\text{head}}\mathbf{\text{labels}}$ and infrequent, low-coverage $\mathbf{\text{tail}}\mathbf{\text{labels}}$, complicating the task of balancing effectiveness across all labels. To address this, combining predictions from multiple retrieval methods, such as sparse retrievers (e.g., BM25) and dense retrievers (e.g., fine-tuned BERT), offers a promising solution. The fusion of $\mathit{\text{sparse}}$ and $\mathit{\text{dense}}$ retrievers is motivated by the complementary ranking characteristics of these methods. Sparse retrievers compute relevance scores based on high-dimensional, bag-of-words representations, while dense retrievers utilize approximate nearest neighbor (ANN) algorithms on dense text and label embeddings within a shared embedding space. Rank-based fusion algorithms leverage these differences by combining the precise matching capabilities of sparse retrievers with the semantic richness of dense retrievers, thereby producing a final ranking that improves the effectiveness across both head and tail labels.