Paul Rodrigues
Ph.D Candidate,
Computational Linguistics Lab

Department of Linguistics
Indiana University


Center for Advanced Study of Language
University of Maryland






New Stuff:
  • 08/07/10 - Defended my dissertation proposal, "Parsing Grammar Variation."
  • 07/31/10 - "Spell Correction for Dialectal Arabic Dictionary Lookup" was accepted to ACM Transactions on Asian Information Processing.
  • 05/19/10 - "Error Correction for Arabic Dictionary Lookup" was presented at LREC 2010 in Malta.
  • 05/17/10 - "Creating a dual-use pandialectal Pashto grammar" was presented at Language Education and Resource Network (LEARN) for AF-PAK Languages. Omaha, Nebraska. May 2010
  • 04/22/10 - Presented "Effects of listener experience with foreign accent on perception of accentedness and speaker age" at ASA 2010 in Baltimore.
  • 09/19/09 - Presented "Determining L1 & L1 Degree of Phonological Accent From Phonetic Transcription after Categorical Filtration of Speech." at TSLL 2009
  • 09/02/09 - Joined the University of Maryland faculty
  • 08/25/09 - Nominated to candidacy at Indiana University
I am a Ph.D candidate in Computational Linguistics at Indiana University, researching speaker variation & categorization, as well as syntactic and phonological acquisition & parsing. I recently joined the research faculty at University of Maryland.

Computational Processing of Linguistic Variation

I am interested in computational processing of linguistic variation (e.g. dialects, accents, and individual differences). I design language processing systems capable of identifying these variations and adapting to them. Recently, I have been working with the processing of bilingual codeswitching between English and dialectal Arabic in chat text.

Linguistic Anomaly Detection

Over the past couple of years, I've been working on systems that detect anomalies in natural language.



I study some Central Asian languages. I posted some flashcards, pronunciation and grammar references for Arabic, Turkmen, and Uzbek--just for fun. They seem to be helpful to a number of people, so I'm leaving them up.