Using Maximum Entropy for Text Classi cation - …
Using Maximum Entropy for Text Classi cationKamal La ertyyla of Computer ScienceCarnegie Mellon UniversityPittsburgh, PA 15213zJust Research4616 Henry StreetPittsburgh, PA 15213AbstractThis paper proposes the use of Maximum en-tropy techniques for text Classi cation . Maxi-mum Entropy is a probability distribution esti-mation technique widely used for a variety ofnatural language tasks, such as language mod-eling, part-of-speech tagging, and text segmen-tation. The underlying principle of maximumentropy is that without external knowledge,one should prefer distributions that are uni-form. Constraints on the distribution, derivedfrom labeled training data, inform the tech-nique where to be minimally non-uniform. Themaximum Entropy formulation has a unique so-lution which can be found by the improved it-erative scaling algorithm. In this paper, max-imum Entropy is used for text Classi cation byestimating the conditional distribution of theclass variable given the document.
Using Maximum Entropy for Text Classi cation Kamal Nigamy knigam@cs.cmu.edu John La ertyy la erty@cs.cmu.edu Andrew McCallumzy mccallum@justresearch.com
Download Using Maximum Entropy for Text Classi cation - …
Information
Domain:
Source:
Link to this page:
Please notify us if you found a problem with this document: