Jacob Eisenstein

Selected Recent Publications

Bayesian Unsupervised Topic Segmentation. Eisenstein and Barzilay. EMNLP 2008.: A new method to segment text and speech transcripts into topically-coherent units, using both lexical cohesion and cue phrases. First paper to show that cue phrases can be used without supervision. combining them with cohesion in a generative Bayesian framework.

Gestural Cohesion for Topic Segmentation. Eisenstein, Barzilay, and Davis. ACL 2008.: Coherent discourse topics contain internally consistent gestural forms, paralleling similar work on the distribution of lexical items. Automatically extracted gesture features are used to improve unsupervised topic segmentation on dialogues.; Unsupervised part-of-speech tagging works better when applied to multiple languages simultaneously.

Modeling Gesture Salience as a Hidden Variable for Coreference Resolution and Keyframe Extraction. Eisenstein, Barzilay, and Davis. Journal of Artificial Intelligence Research, volume 31, 353-398.: Describes the use of a conditional hidden variable model for gesture salience in coreference resolution. The model improves performance on coreference resolution, and the estimates of gesture salience can be transferred directly to select keyframes containing interesting gestures.; A quantitative analysis of the influence of speaker and topic on gestural form. Using low-level interest point features, it is possible to show that multiple speakers use similar gestures when describing the same topic.

Jacob Eisenstein

Selected Recent Publications

Software

Election Prediction

Contact

		Jacob Eisenstein I'm a Postdoctoral Fellow at the Beckman Institute of the University of Illinois. My primary research interest is learning computational models for coverbal gestural communication. I use machine learning to identify patterns of gesture that can be leveraged to improve performance on natural language understanding tasks. I also work on structured learning techniques for language processing, especially hierarchical Bayesian models. Other research interests include intelligent and multimodal user interfaces. publications \| dissertation \| for non-specialists
Selected Recent Publications Bayesian Unsupervised Topic Segmentation. Eisenstein and Barzilay. EMNLP 2008. A new method to segment text and speech transcripts into topically-coherent units, using both lexical cohesion and cue phrases. First paper to show that cue phrases can be used without supervision. combining them with cohesion in a generative Bayesian framework. Gestural Cohesion for Topic Segmentation. Eisenstein, Barzilay, and Davis. ACL 2008. Coherent discourse topics contain internally consistent gestural forms, paralleling similar work on the distribution of lexical items. Automatically extracted gesture features are used to improve unsupervised topic segmentation on dialogues. Unsupervised Multilingual Learning for POS Tagging. Snyder, Naseem, Eisenstein and Barzilay. EMNLP 2008. Unsupervised part-of-speech tagging works better when applied to multiple languages simultaneously. Modeling Gesture Salience as a Hidden Variable for Coreference Resolution and Keyframe Extraction. Eisenstein, Barzilay, and Davis. Journal of Artificial Intelligence Research, volume 31, 353-398. Describes the use of a conditional hidden variable model for gesture salience in coreference resolution. The model improves performance on coreference resolution, and the estimates of gesture salience can be transferred directly to select keyframes containing interesting gestures. Discourse Topic and Gestural Form. Eisenstein, Barzilay, and Davis. AAAI 2008. A quantitative analysis of the influence of speaker and topic on gestural form. Using low-level interest point features, it is possible to show that multiple speakers use similar gestures when describing the same topic. Software Some code for machine learning, computer vision, video annotation, and other random stuff. Election Prediction A side interest of mine is statistical models for predicting election results based on polls. I did pretty well predicting the US Senate elections in 2006. Contact Beckman Institute for Advanced Science and Technology University of Illinois Urbana-Champaign 405 N. Mathews St. Urbana, IL 61801 617-253-2663 jacobe gmail com