Zvika Marx - the Home Page Starting August 2006: Research Scientist, shopping.com (an eBay company), Netanya, Israel Here I work on automated and semi-automated creation of product catalogs for e-commerce February 2005 -- July 2006: Postdoc, CS & AI Lab, Massachusetts Institute of Technology I worked on developing methods for transfer learning. The idea was to build a system that can learn to perform a task in one domain, and then transfer some of what it has learned to improve its performance in a new domain. This task is part of CALO project. I worked on it together with Mike Rosenstein, supervised by Leslie Kaelbling and Tom Dietterich from Oregon State University. February 2004 -- January 2005: Postdoc, Proteus Project, New York University The focus of my research in NYU has been developing semi-supervised learning methods and applying them for Named Entity recognition. My supervisor was Satoshi Sekine. For several years prior to that I was a PhD Student in the Neural Computation PhD Program, The Hebrew University of Jerusalem (supervisors: Eli Shamir, Ido Dagan) e-mail: marxzv / gmail

  Publications

• PhD thesis:
    Structure Based Computational Aspects of Similarity and Analogy in Natural Language
abstract (text), dissertation (pdf), chapters (links to pdf files).
 
• Transfer learning with an ensemble of background tasks
with Michael T. Rosenstein, Leslie Pack Kaelbling and Thomas G. Dietterich (pdf).
NIPS 2005 Workshop on Inductive Transfer: 10 Years Later.
 
• A generalized framework for revealing analogous themes across related topics
with Ido Dagan and Eli Shamir (pdf).
HLT/EMNLP 2005: Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing, (pp. 979-986).
 
• Identifying structure across pre-partitioned data
with Ido Dagan and Eli Shamir (gzipped ps, pdf).
NIPS*03: Seventeenth Annual Conference on Neural Information Processing Systems, (pp. 489-496).
 
• Clustering polysemic subcategorization frame distributions semantically
with Anna Korhonen and Yuval Krymolowski (pdf).
ACL 2003: 41st Annual Meeting of the Association for Computational Linguistics, (pp. 64-71).
 
• Cross-dataset clustering: Revealing corresponding themes across multiple corpora
with Ido Dagan and Eli Shamir (pdf).
CoNLL-2002: Sixth Conference on Natural Language Learning, (pp. 15-21).
 
• Cross-component clustering for template induction
with Ido Dagan and Eli Shamir (pdf).
Workshop on Text Learning (TextML-2002), held at ICML-2002, (pp. 66-75).
 
• Coupled clustering: a method for detecting structural correspondence
  
Full version, with Ido Dagan, Joachim M. Buhmann and Eli Shamir (gzipped ps, pdf).
Journal of Machine Learning Research, (2002, pp. 747-780).
  An earlier version, with Ido Dagan and Joachim M. Buhmann appeared in ICML 2001, (pp. 353-360).
 
• Detecting sub-topic correspondence through bipartite term clustering
with Ido Dagan and Eli Shamir, (pdf).
ACL-99 Workshop on Unsupervised Learning in Natural Language Processing, (pp. 45-51)
 
• An abstract of my PhD research proposal describing in short what I had in mind then.