Projekt
corpus_pepys: The Diary of Samuel Pepys — Interactive Semantic Knowledge Graph
corpus_pepys turns the complete diary of Samuel Pepys (1660–1669) into a standalone, queryable semantic knowledge graph. The 3,355 daily entries are parsed, NLP-enriched and topic-classified into 7,285 chunks, then indexed as a DiaryKG knowledge graph — 41,517 nodes, 333,679 edges — with a sqlite-vec vector store for…
corpus_pepys turns the complete diary of Samuel Pepys (1660–1669) into a standalone, queryable semantic knowledge graph. The 3,355 daily entries are parsed, NLP-enriched and topic-classified into 7,285 chunks, then indexed as a DiaryKG knowledge graph — 41,517 nodes, 333,679 edges — with a sqlite-vec vector store for sub-second natural-language retrieval. The index is baked into a self-contained Docker image and served through the KGRAG federated query layer, requiring no index build, model download or network access at run time.