FrameNet

FrameNet is an ongoing corpus-based computational lexicography research and resource development project based on the principles of Frame Semantics (Fillmore 1975, 1977, 1982, 1985). FrameNet documents the valences, or semantic and syntactic combinatorial possibilities, of English vocabulary, providing a unique training dataset for automatic semantic role labeling in natural language processing applications, such as event recognition, machine translation and sentiment analysis. For detailed information about FrameNet, its history, current status, and products, please see the website.