Bevan K. Jones, project researcher
I am a project researcher at AIST (Tokyo, Japan) working in the Artificial Intelligence Research Center.
I recently completed a postdoc at the National Institute of Informatics working with Yusuke Miyao after completing my PhD with my advisors Sharon Goldwater and Mark Johnson.
My research interests center on natural language understanding, with a focus on grammar induction and how sentence meanings can influence the grammars one learns.
I approach the problem as an information theoretic and computational problem, leveraging tools from computer science and mathematics to explore questions relevant to a range of problems in psychology and natural language technology.
I do a lot of work in semantic parsing, developing systems that automatically learn to map from sentences of English or other natural languages like German, Turkish, or Japanese to formal representations of their meanings.
I have also worked on related problems such as semantics-based statistical machine translation, and on jointly learning word-segmentations and word meanings.
I often use Bayesian inference techniques, applying them to models taken from formal language theory such as graph grammars and tree transducers, and lean toward machine learning techniques involving distant supervision.
Bevan K. Jones, Sharon Goldwater, and Mark Johnson (2013). Modeling graph languages with grammars extracted via tree decompositions. In Proceedings of the 11th International Conference on Finite-State Methods and Natural Language Processing (FSMNLP2013).
David Chiang, Jacob Andreas, Daniel Bauer, Karl Moritz Hermann, Bevan Jones, and Kevin Knight (2013). Parsing graphs with hyperedge replacement grammars. ACL 2013.
Bevan K. Jones*, Jacob Andreas*, Daniel Bauer*, Karl Moritz Hermann*, and Kevin Knight (2012). Semantics-Based Machine Translation with Hyperedge Replacement Grammars. In Proceedings of the 24th International Conference on Computational Linguistics (COLING2012).
* All four first authors contributed equally. Their order is random.
Bevan K. Jones, Mark Johnson, and Sharon Goldwater (2012). Semantic Parsing with Bayesian Tree Transducers. In Proceedings of the 50th Annual Meeting of the Association of Computational Linguistics (ACL2012). (Data set.)
Bevan K. Jones, Mark Johnson, and Sharon Goldwater (2011). Formalizing Semantic Parsing with Tree Transducers. The 2011 Australasian Language Technology Workshop (ALTA2011).
Benjamin Börschinger, Bevan K. Jones, and Mark Johnson (2011). Reducing Grounded Learning Tasks To Grammatical Inference. In Proceedings of the 2011 Conference on Empirical Methods in Natural Language Processing (EMNLP2011), pages 1416-1425.
Mark Johnson, Katherine Demuth, Michael C. Frank, and Bevan K. Jones (2010). Synergies in learning words and their meanings. Neural Information Processing Systems (NIPS2010).
Bevan K. Jones (2010). A Distributional Account of Covariance Effects and Talker Adaptation in Infant and Adult Phonetic Category Recognition. In Proceedings of the 32nd Annual Meeting of the Cognitive Science Society.
Bevan K. Jones, Mark Johnson, and Michael C. Frank (2010). Learning Words and Their Meanings from Unsegmented Child-directed Speech. Human Language Technologies: The 2010 Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL2010), pp. 501-509.
Affiliations (Past & Present)