Working at MIT offers opportunities, an environment, a culture – and benefits – that just aren’t found together anywhere else. If you’re curious, motivated, want to be part of a unique community, and help shape the future – then take a look at this opportunity.
POSTDOCTORAL ASSOCIATE-LOW-RESOURCE LANGUAGE LEARNING, Brain and Cognitive Sciences (BCS), to join a large multidisciplinary, multi-PI project on low-resource language learning at the intersection of natural language processing, machine learning, linguistics, and cognitive science. Human language technology has seen remarkable advances in the past several years, but relies heavily on data-hungry deep-learning models. The huge datasets required for high-performance levels with these technologies remain a privilege for only a handful of the world’s 6,000+ languages, keeping the fruits of NLP and machine learning from reaching everyone on the planet--especially the Global South. For linguistics and cognitive science, understanding of language structure, learning, and processing is similarly skewed toward resource-rich languages. Will develop and test machine learning methods for natural language processing that perform well for typologically-diverse languages under data-scarce conditions and that help address questions of scientific interest in linguistics and cognitive science; and develop and maintain an independent research program in synergy with project goals. Potential topics include automatic discovery of compositional morphological, syntactic, and semantic structure; induction of phonological rule or constraint systems; interactive, grounded learning of linguistic structure and meaning; and phylogenetic and areal modeling of language change. The project is supported by the MIT–IBM Watson AI Laboratory with direct supervision by Roger Levy. Other PIs include Jacob Andreas, Regina Barzilay, Shiyu Chang, Jim Glass, Tommi Jaakola, Yoon Kim, Tahira Naseem, Mo Yu, and Yang Zhang.
REQUIRED: Ph.D. in computer science, linguistics, cognitive science, or related field by the start date; expertise with natural language processing, deep learning, linguistic theory, and Bayesian statistics; and a background/interest in understudied languages, linguistic fieldwork language documentation, and linguistic typology. Job #19430
Target start date is summer or fall 2021, with some potential flexibility.
Questions or inquiries about the position may be directed to Roger Levy at firstname.lastname@example.org.
In addition to applying via the MIT website with a CV, cover letter, research statement, and contact information for three potential letter writers (which may be included as a single pdf on the MIT site), applicants are asked to submit their material to https://academicjobsonline.org/ajo/jobs/18385.
To ensure fullest consideration of your application, please apply by April 16, 2021.