Large language models struggle to analyze complex documents such as academic papers and financial reports because existing approaches flatten documents into plain text, discarding structural information. DocMaster parses documents into hierarchical document trees that preserve original layouts and constructs a structure-aware semantic index enabling accurate document filtering and in-depth analysis, with an interactive interface for uploading document collections, building semantic indices, filtering via natural-language queries, and question answering over filtered results.
@article{chen2026docmaster,title={DocMaster: A Hierarchical Structure-Aware System for Document Analysis},author={Chen, Ziqi and Zhou, Yingli and Zhang, Fangyuan and Xu, Quanqing and Yang, Chuanhui and Fang, Yixiang},journal={arXiv preprint arXiv:2607.08539},year={2026},month=jul,archiveprefix={arXiv},primaryclass={cs.CL},url={https://arxiv.org/abs/2607.08539},}
