(1) The Challenge: Isolated Knowledge & the Need for Grounding

Many data science and knowledge-based projects naturally involve creating custom schemas or lightweight ontologies – defining terms like PROJECT_LEAD, SALES_REGION, or COMPONENT_FAILURE_MODE that make sense within a specific context (e.g., in simple YAML files). While valuable locally, these schemas often become isolated knowledge silos. Without alignment to a common standard, integrating data across projects becomes difficult, querying consistently is a challenge, and opportunities to leverage vast amounts of external, structured knowledge are missed. How can we connect these custom concepts to the wider world?

This is where Wikidata comes in. Why map to Wikidata specifically?

• It's one of the largest, collaboratively edited, multilingual knowledge bases available, acting as a central structured data hub for Wikimedia projects and countless external applications ().
• Its vast scale and broad domain coverage (from proteins to companies to historical events) make it an ideal universal anchor point for grounding diverse custom schemas.
• Mapping provides immediate benefits:
  • Standardization: Your SALES_REGION aligns with a standard Wikidata geo-identifier (QID).
  • Interoperability: Your graph can now connect with other datasets linked via Wikidata IDs.
  • Powerful Data Enrichment: Link your internal 'Acme Corp' entity to its Wikidata QID, and you can instantly query for its industry (P452), headquarters coordinates (P625), official website (P856), subsidiaries (P355), and much more, automatically enriching your own data.

However, the mapping process itself is a major bottleneck. Manually finding the correct Wikidata QID/PID for potentially hundreds of custom terms is extremely tedious, requires niche expertise, and is error-prone. Fully automated alignment tools often struggle with the semantic ambiguity inherent in custom schemas (e.g., does 'position' mean job title P39 or coordinate P625?).

(2) Our Approach: LLM-Assisted Interactive Mapping

This talk introduces a practical Python tool and workflow designed to dramatically accelerate the process of mapping custom schema elements (from formats like YAML) to Wikidata identifiers, while maintaining high accuracy through efficient human oversight. We combine the semantic reasoning of Large Language Models (LLMs) with an interactive workflow.

(3) Workflow Walkthrough:

We will walk through the tool's workflow:

• Input: Starting with a custom schema/ontology (e.g., a YAML file defining entity types like 'PERSON', 'COMPANY' and relations like 'worksFor').
• Candidate Search: Automatically searching Wikidata (API client with caching and retries) for potential QID/PID matches.
• LLM Suggestion: Using dspy to prompt an LLM with source term details and candidates, generating structured JSON suggestions (LLMMappingSuggestion Pydantic model) with confidence scores and reasoning.
• Interactive Review: Demonstrating the inquirer-based CLI for quickly accepting/overriding/skipping LLM suggestions, aided by confidence thresholding for auto-acceptance.
• Batch Modes: Explaining the scalable --batch-suggest and --batch-review modes for handling larger schemas efficiently.
• Hierarchy Suggestion: Showcasing how the tool uses Wikidata's parent information (via SPARQL and another LLM prompt) to interactively suggest adding subclass_of links to the user's own schema, helping refine its structure.
• Output: Generating the final schema file (YAML) enriched with Wikidata mappings and confirmed hierarchy links.

(4) Technical & Engineering Highlights:

• Briefly mention the Python libraries (dspy, requests, tenacity, inquirer, pydantic, platformdirs, yaml).
• Highlight robustness features: API retries, caching, structured error handling, decoupled design.

(5) Demo: (Mention plan for a brief demo showcasing the interactive mapping and hierarchy suggestion.)

(6) Benefits & Conclusion:

This approach provides a pragmatic solution to a common data integration bottleneck. By intelligently combining LLM suggestions with efficient human validation, it allows teams to:

• Significantly speed up mapping custom schemas to Wikidata.

• Improve mapping consistency and accuracy.

• Enrich internal knowledge graphs by linking them to a global standard.

• Leverage Wikidata's structure to refine custom schemas.

Attendees will learn about a practical workflow applying LLMs to KG alignment and gain insights into building robust, interactive data tools with the modern Python stack.