Report on Matching and Clustering Entities in Large Collections of Encoded Archival Context

I will be reporting my progress towards implementing techniques that match and merge entities in collections of Encoded Archival Context (Corporate Names, Persons and Families) records. I would be discussing cases where our initial simple techniques, techniques based on exact matches using name authority files as a reference, failed to identify matches. I also plan to discuss my experiments on using probabilistic graphical models to cluster entities based on the information present in these records.