Case Study: Los Alamos National Labs achieves 99% name-disambiguation accuracy across terabytes with Franz Inc.'s AllegroGraph and Hadoop

A Franz Inc. Case Study

Preview of the Los Alamos National Labs Case Study

Los Alamos National Labs - Customer Case Study

Los Alamos National Labs faced a Big Data name-disambiguation challenge: processing terabytes of structured and semi-structured bibliographic metadata to identify people, co-authors, affiliations and hidden relationships despite spelling variants, nicknames and abbreviations. To solve this, Los Alamos used Franz Inc.'s AllegroGraph semantic graph platform integrated with a Hadoop-based processing stack.

Franz Inc. implemented AllegroGraph alongside Hadoop (HDFS, MapReduce), streaming input and Mahout for metadata extraction and machine learning, creating RDF triples and ontologies (e.g., Dublin Core, FOAF) for threshold-based matching and network analysis. The combined solution delivered 99% accuracy in identifying and disambiguating people across terabyte-scale datasets, improved discovery of affiliations and connectedness, and a scalable architecture that grows with real-world needs.


Open case study document...

Franz Inc.

59 Case Studies