DOSA: A Tree-Guided, Self-Regressive Framework for Long Document Structure Analysis
For researchers in document understanding and information retrieval, this work addresses the challenge of long-range dependencies in multi-page documents with a novel framework that leverages structural context, showing strong gains on a challenging benchmark.
The paper proposes DOSA, a tree-guided self-regressive framework for document structure analysis that processes documents chunk-by-chunk to infer hierarchical and ordering relations among page objects. On the DocHieNet benchmark, DOSA achieves improvements of up to 4 F1 points and 19 TEDS points.
In visually-rich documents, information is encoded not only in individual page objects such as tables, headers, and text blocks, but also in the structural relations among them, making document structure analysis fundamental to information retrieval and document understanding. However, accurately inferring such relations remains challenging in multi-page documents with long-range dependencies and heterogeneous layouts. To address this, we propose a tree-guided and self-regressive framework, termed DOcument Structure Analyzer (DOSA), for inferring relations among page objects and reconstructing document-level semantic trees. DOSA processes documents chunk-by-chunk, fusing visual, textual, and layout features for each page object and predicting hierarchical and ordering relations. The predicted relations are used to incrementally construct a semantic tree, which is then leveraged as structural context to guide inference on subsequent chunks. Experimental results on five benchmarks demonstrate the effectiveness of DOSA, with improvements of up to 4 F1 points and 19 TEDS points on DocHieNet, the most challenging multi-page hierarchy benchmark.