Class Meeting for 10-707 10/6/2010

This is one of the class meetings on the schedule for the course Information Extraction 10-707 in Fall 2010.

Ranking Perceptrons for NER

• Slides
• Notes (these are the same as for Monday's class)

Required Readings

• Discriminative Training Methods for Hidden Markov Models: Theory and Experiments with Perceptron Algorithms, Collins, EMNLP 2002.

Optional Readings

• Large Margin Methods for Structured and Interdependent Output Variables, Tsochantaridis et al, JMLR 2005. A boosting approach to structured prediction.
• Exponentiated gradient algorithms for log-linear structured prediction, by A. Globerson, T. Y Koo, X. Carreras, M. Collins. In Proceedings of the 24th international conference on Machine learning, 2007.. A more recent EG-based approach.
• Phrase clustering for discriminative learning, by D. Lin, X. Wu. In Proceedings of the Annual Meeting of the Association for Computational Linguistics and the International Joint Conference on Natural Language Processing of the Asian Federation of Natural Language Processing, 2009.. Using large-scale unsupervised learning to classify search-query phrases into types (e.g., person, place, ...). Somewhat like the Collins & Singer paper in aims, but using modern corpora.

Background Readings

• William W. Cohen & Sunita Sarawagi (2004): Exploiting Dictionaries in Named Entity Extraction: Combining Semi-Markov Extraction Processes and Data Integration Methods in KDD 2004: 89-98.