|
SEMINARMachine Learning for Text Classification and Information Organization
Half Day, Sunday, November 17, 1:30pm - 5:00pm (separate fee) This half-day tutorial shows how to use machine learning in categorizing, filtering, mining, and otherwise imposing sense on online data. By automating not just the classification of text,
but the production of the classification rules themselves, machine learning can make classification practical in situations where purely manual indexing is too slow or costly.You will learn the major approaches to training text
classification systems, measuring and enhancing their effectiveness, and combining them with manual classification in the most economical fashion. Issues discussed will include evaluating commercial text classification offerings,
embedding text classification in larger systems, and making "build vs. buy" decisions. Concepts will be illustrated with examples of operational systems from domains such as knowledge management, customer service
automation, alerting systems, web directories, vertical portals, category-oriented search, spam and porn filtering, text data mining, and survey research.Instructor Dave Lewis
David Lewis is an independent consultant (
www.DavidDLewis.com) based in Chicago, IL. He advises clients on the development and effective use of technology for classifying, mining, and retrieving
text, as well as related areas of natural language processing, machine learning, and bioinformatics. His clients have included companies of all sizes, as well as universities and research consortia. Prior to launching his
consulting practice, he was a researcher for eight years at Bell Labs and AT&T Labs, and before that a faculty member at the University of Chicago. He has authored over 40 technical papers on information retrieval and machine
learning, has 5 patents, and has been extensively involved in the design of U.S. Government evaluations of information retrieval and natural language processing technology. |
