April 24, 2013
Relation extraction from lobbyist disclosure forms. Parsing legislative text and the Congressional Record to attempt information extraction and cluster similar speakers/authors/documents.
Parallel processing, deep learning, building searchable parsed corpora, working with limited-resource languages.
Everything that I have so far is still in early development stages, but I can speak to NLP-relevant resources offered through the Sunlight Foundations APIs...
Developer at the Sunlight Foundation, studied Linguistics/NLP at UPenn.