CDM learning algorithm : an approach to learning for text caterorization /
There have been two primary approaches to the problem of text categorization, the knowledge engineering approach, and the machine learning approach. This research focuses on the latter. The motivation is that statistical problems have been associated with natural language text when it is applied a...
| Main Author: | |
|---|---|
| Format: | Thesis Book |
| Language: | English |
| Published: |
[Place of publication not identified] :
[publisher not identified] ;
1996.
|
| Subjects: | |
| Online Access: | http://proxy.library.tamu.edu/login?url=http://proquest.umi.com/pqdweb?did=739363551&sid=1&Fmt=2&clientId=2945&RQT=309&VName=PQD |
| Summary: | There have been two primary approaches to the problem of text categorization, the knowledge engineering approach, and the machine learning approach. This research focuses on the latter. The motivation is that statistical problems have been associated with natural language text when it is applied as input to existing machine learning algorithms (too much noise, too many features, skewed distribution) . The Category Discrimination Method (CDM) addresses these statistical problems; it is designed specifically for text categorization and has biases that may be more appropriate for the domain of natural language text. The bases of the CDM are results from cognitive psychology about the way that humans learn categories and concepts vis-Þ-vis contrasting concepts. The essential formula is cue validity, which is used to select from all possible single word-based features the best predictors of a given category. The hypothesis that the CDM's performance will exceed two non-domain specific algorithms, Bayesian classification and decision tree learners, is empirically tested. |
|---|---|
| Item Description: | Vita. "Major Subject: Computer Science". |
| Physical Description: | x,172 leaves : illustrations ; 28 cm. Issued also on microfiche from University Microfilms Inc. |
| Bibliography: | Includes bibliographical references: pages 102-107. |