ARABIC TEXT MINING AND ROUGH SET THEORY FOR DECISION SUPPORT SYSTEM

Abstract

The present work based on an implement classification system for Arab Complaint. The
basis of this system is the construction of Decision Support Systems, Text Mining, and Rough Set, a
specialized semantic approach of correlation which is referred to as the model of ASN for creating
semantic structure between all document’s words. A particular semantic weight represents true
significance of this term in the text after each feature is probable to be utilized the semantic selection
of a suggested feature based on 2 threshold values, the first being the maximum weight in a document
and the other one representing the characteristic having the maximum semantic scale with the first one
of the thresholds. The task of classification in the stage of testing depends on the dependency degree
concept in the raw group theory than before to treat all the characteristics of every one of the classes
that results from the stage of training as particular condition rules and therefore, considered in the
lower rounding group. The outputs of this classification system are sufficient, performance is quite
good persuasive, and assessment of this the system through measurement precision.