baojie/shiji-kb

Shiji-KB is a knowledge graph platform that transforms classical Chinese historical texts into interactive, explorable digital resources.

The project addresses the challenge of making ancient texts computationally accessible and analytically useful. It converts the Records of the Grand Historian, a foundational 570,000-character historical work, into a structured knowledge graph using AI-powered entity recognition and relationship extraction. The result is presented through an interactive wiki interface where users can navigate person-to-person relationships, trace event timelines, search across annotated text, and follow hyperlinked references.

The tool suits historians, academic researchers, and content creators seeking to work with classical Chinese sources. Historians gain syntax-highlighted reading with one-click navigation and relationship visualization. Researchers benefit from contradiction detection, source tracing, and systematic reasoning capabilities built into the knowledge structure. Content creators can use it as a historical reference engine and material library. The project includes methodological documentation covering both conceptual frameworks and implementation pipelines, making it relevant for those studying knowledge engineering approaches applied to historical texts.Development activity shows sustained expansion of the knowledge base through automated quality pipelines that systematically upgrade stub entries to featured and premium status. The project maintains an interactive timeline system styled as historical transit maps and has integrated historical cartography by automatically matching locations to period-appropriate map excerpts. A butler system has executed thousands of maintenance operations including new page creation, content enrichment, citation verification, and broken link repair.