Brian Roark

• Research Interests

Syntactic parsing of text and speech; language modeling for automatic speech recognition; supervised and unsupervised learning of language and parsing models; probabilistic models of human language processing; spoken language processing for diagnosis of neurodevelopmental and neurodegenerative disorders


• Software
• Grammar Library (GRM Library): general software collection for constructing and modifying weighted automata and transducers representing grammars or language models.
• SParseval scoring tool for evaluating parse accuracy when parsing ASR output. Visit this page for technical details and software download.
• Selected Grants
• Translational refinement of adaptive communication system for locked-in patients,
  NIH, NIDCD Grant (PI Fried-Oken), 2009-2014
• Efficient hidden structure annotation via structural multiple-sequence alignments,
  NSF (IIS) RI Grant, 2008-2011
• DHB: Measuring Spoken Language Variability in Elderly Individuals,
  NSF (BCS) HSD Grant, 2008-2011
• Discriminative Syntactic Language Modeling: Automatic Feature Selection and Efficient Annotation,
  NSF (IIS) Career Grant, 2005-2010
• Automated Analysis of Spoken Story Recall Tests,
  Oregon Center for Aging and Technology Pilot Grant, 2005-2006.
• Selected Publications
  
• Curriculum Vitae
  
• Participated in JHU Summer Workshop, 2005.
  
• Member, IEEE Speech and Language Technical Committee
  
• Students:
  • Nathan Bodenstab
  • Yongshun Chen
  • Aaron Dunlop
  • Seeger Fisher
  • Kristy Hollingshead