DocMaster: A Hierarchical Structure-Aware System for Document Analysis
2026-07-09 • Databases
DatabasesArtificial Intelligence
AI summaryⓘ
The authors created DocMaster, a system that helps computers better understand complex documents like academic papers by keeping their original structure (such as sections and tables) instead of turning everything into plain text. This structure-aware approach improves how well the system finds relevant documents and answers questions about them. Users can upload documents, search through them using natural language, and get detailed answers within an interactive web interface. The authors also made their code and demo available online.
large language modelsdocument analysishierarchical structuresemantic indexingnatural language processingquestion answeringdocument filteringinteractive web interface
Authors
Ziqi Chen, Yingli Zhou, Fangyuan Zhang, Quanqing Xu, Chuanhui Yang, Yixiang Fang
Abstract
Leveraging large language models (LLMs) to analyze complex documents -- such as academic papers, technical manuals, and financial reports -- has emerged as a mainstream and critical task in both research and industry. In practice, users must first filter relevant documents from large collections and then conduct in-depth analysis (e.g. question answering) over the selected subset, yet existing systems flatten documents into plain-text chunks, discarding the rich hierarchical structures (sections, tables, figures, equations) and degrading downstream performance. We present DocMaster, a hierarchical structure-aware document analysis system. DocMaster parses documents into hierarchical document trees preserving original layouts and constructs a structure-aware semantic index that enables accurate document filtering and in-depth analysis. We demonstrate DocMaster through an interactive web interface that enables users to upload document collections, construct tree-based and multi-view semantic indices, filter relevant documents via natural-language conditions, and perform follow-up question answering over the filtered results. The source code, data, and demo are available at https://doc-master.github.io/.