In the era of Agentic AI, open historical knowledge bases like Wikidata and DBpedia often fall short due to severe data fragmentation. Every developer aiming to map complex, deeply interconnected lineages must understand how to handle compounding data failures like sparse genealogical properties, cross-source entity duplication, encoding corruption, and rigid retrieval limitations for dynamic relational queries.
In this session, Annisa and Bara will explore a two-phase architecture designed to cure corrupted historical open data and turn it into a live intelligence layer. They will break down a modular five-stage data enrichment pipeline that cleanses, deduplicates, and loads open data into Neo4j , followed by the deployment of a Model Context Protocol (MCP) server that exposes the graph database as a live tool suite for an autonomous LLM agent. You will learn how to implement fuzzy entity disambiguation across heterogeneous sources, apply Adamic-Adar graph analytics to mitigate metric distortion from empty nodes, and expose Neo4j tools via MCP for dynamic Cypher execution. You will see a demonstration of how AI-powered imputation successfully recovered 88% of missing records for 108 historical figures using auditable confidence scoring, along with lessons learned in automated data cleansing and agentic workflows. Join this session to get a replicable blueprint for transforming messy, fragmented open data into an interactive, graph-verified intelligence layer.
Speaker
Annisa Nur Fauzi
Information Systems Student, Institut Teknologi Sepuluh Nopember (ITS)
Annisa Nur Fauzi is a 6th-semester Information Systems Student at Institut Teknologi Sepuluh Nopember (ITS) who considers herself a passionate, lifelong learner. Embracing a coder-at-heart vibe, she genuinely loves digging deep into messy data to unlock hidden connections. Annisa has a broad curiosity that spans data engineering, historical lineages, and environmental insights. As an evolving student researcher, she is always eager to experiment with new tools, exploring how graph technologies and AI agents can solve data challenges while constantly expanding her technical horizons.
