The Effectiveness Of Feature Selection Metrics On Text Categorization.

      Аннотация: The human experts usually do the documents cataloguing and indexing manually. With the growth of online information, and sudden expansion in the numerous electronic documents provided on the web and digital libraries, there is difficulty in categorizing in both electronic documents and traditional library materials using only a manual approach. To solve these problems as well as improve the efficiency and effectiveness of document categorization at the library setting. However, the main idea of text categorization is to allot textual documents data according to one or more predetermined topic codes on the basis of knowledge accumulated in the training process. Finally, text-categorization is a very good technique that uses labeled training data for learning the classification system, such as (BayesNet, NaiveBayes, Trees, etc.) and then automatically classes the remaining text by applying the learned system. For instance, it can determine the words such as GalaxyS5", GalaxyS4, or Note5" those are related to category of technology named Samsung, and then their documents must belong to that category.

Ключевые слова: Learning Machines, text categorization, Feature Selection Metrics