Current Research Projects

Predicting Secondary Events (Consolidated Edison Company of New York)

Loqui: a Wizard Ablation Dialog System Project (NSF) [ppt]

• Transcription guidelines: for transcribing a corpus of patron/librarian phone calls

MASC: Manually Annotated Subcorpus of the American National Corpus (NSF)

• Heterogeneous corpus of current American English: multiple modalities, genres
• Validation of automatic annotation of NP Chunks, VP Chunks, Named Entities
• Word sense disambiguation
• Interannotator Agreement

Dialog Function Units (DFUs; Johns Hopkins Center of Excellence)

• Development of DFU annotation
• Annotation Manual: Version 2.0
Annotators must use Version 2.0. This supercedes earlier versions.
• Testing manual annotation of multiple types of interaction
• Machine learning of automated annotation