Text Classification Version 1.0
STATUS KEY : Specification Design and Architecture Development and Testing Complete
Classifies a body of text (for use in spam detection software).
FunctionalityThe Text Classification component provides a means to execute standard text classification algorithms against bodies of text. The component allows for different algorithm implementations to be added, configured and used. The component comes prepackaged with an implementation of the Naive Bayesian (multi-variate Bernoulli event) algorithm.
A spam detection application could use this component to determine whether or not incoming mail is spam. The component can later be configured to use a different algorithm if the results of the spam detection are unsatisfactory.
Adobe Acrobat is required to view TopCoder Software specification documentation.
Component SpecificationComponent Specification
Requirements SpecificationRequirements Specification
Sequence DiagramClassify Document, General Case
Classify / Reliability, Naive Bayesian
Classify Document, Word Presence
Create Text Classifier
Reliability Measurement, General Case
Set Default Algorithm
Set Default Class Configuration
Set Initial Document Classifications
Use Case DiagramMain