In the era of big data, linking and reconciling information across different datasets is crucial for creating structured, meaningful connections. Developed by UNIMIB as part of the enRichMyData toolbox, LamAPI is designed to simplify this process by enabling efficient entity lookup, data indexing, and enhanced reconciliation across Knowledge Graphs (KGs).
What is LamAPI?
LamAPI helps digest a target data source (e.g., a Knowledge Graph) and prepare it for fast entity lookup and matching operations. While its primary function is to index a KG, it also supports a range of additional features, making it an essential tool for data linking and reconciliation.
Key Features of LamAPI
🔹 Fast entity lookup & indexing – Ensures efficient searching and matching in large-scale KGs like Wikidata and DBpedia.
🔹 Cross-graph lookup – Stores links across different graphs (e.g., sameAsrelations) to enhance reconciliation.
🔹 Retrieval of entity relations & literal values– Extracts valuable context for improved linking.
🔹 Entity embeddings with RDF2vec– Helps estimate the relatedness between candidate matches for more accurate linking.
🔹 Data type identification – Automatically detects specific data types, such as numbers, emails, dates, telephone numbers, street addresses, and URLs.
With these capabilities, LamAPI simplifies the complex process of entity linking, making data integration more seamless and scalable.
What’s New in LamAPI?
Recent updates bring significant enhancements to LamAPI, focusing on improved efficiency, usability, and precision.
✅ Smarter Candidate Filtering
- Entities are now classified into four broad categories: PERSON, LOCATION, ORGANIZATION, and OTHER.
- This functionality improves interactive linking by eliminating unrelated candidates and enhances domain-specific linking, such as company data enrichment (e.g., in the InnoGraph business case).
✅ Enhanced User Control
- Users can now specify the language for entity lookup.
- The newly named entity recognition (NER) type of filtering refines searches.
- Option to include QIDs in responses for easier integration into workflows.
✅ Refactored Code for Simplicity
- The lookup endpoint has been streamlined for better usability.
- Configuring LamAPI for custom Knowledge Graphs is now more intuitive.
Open-Source & Ready to Use
LamAPI is fully open-source, allowing users to download, host, and run it independently. The code for replicating experiments is also available, ensuring transparency and reproducibility for researchers and developers.
Why Use LamAPI?
🔹 Efficient – Speeds up entity lookup and reconciliation. 🔹 Scalable – Works with large datasets like Wikidata & DBpedia. 🔹 Customizable – Can be configured for proprietary KGs. 🔹 User-Friendly – New updates enhance control and interaction.
Try LamAPI Today!
If you’re looking for a powerful, open-source solution to improve data linking and reconciliation, LamAPI is ready to enhance your workflows.