DeepTileBars: Visualizing Term Distribution for Neural Information Retrieval

Most neural Information Retrieval (Neu-IR) models derive query-to-document ranking scores based on term-level matching. Inspired by TileBars, a classic term distribution visualization method, in this paper, we propose a novel Neu-IR model that models query-to-document matching at the subtopic and higher levels. Our system first splits the documents into topical segments, "visualizes" the matching between the query and the segments, and then feeds the interaction matrix into a Neu-IR model, DeepTileBars, to obtain the final ranking score. DeepTileBars models the relevance signals happening at different granularities in a document's topic hierarchy. It thus better captures the discourse structure of the document and the matching patterns. Although its design and implementation are light-weight, DeepTileBars outperforms other state-of-the-art Neu-IR models on benchmark datasets including the Text REtrieval Conference (TREC) 2010-2012 Web Tracks and LETOR 4.0.