An Automatic Text Classification System Based on Genetic Algorithm
Mohammed Khaleel, Ismail Hmeidi, Hassan Najadat · 2016
The increasing numbers of on-line text documents make the process of searching and accessing documents related to a specific category a very difficult task. By classifying the documents, the search is then limited to only those documents that related to a particular category. Text classification is the process of classifying documents based on their content into predefined set of categories. Many classification systems that based on rules generation approach have been adopted for text classification. The classification rules that generated from these classifiers conducted directly from the characteristics of training documents. Which will be limited to a certain categories and has unequal number of the generated classification rules per category. In this paper, an automatic text classification system based on the genetic algorithm classifier has been developed. The genetic algorithm classifier generates a predefined number of optimized classification rules that have high level of flexibility and cover wide range of the characteristics that belong to the training documents. The performance of the genetic algorithm classifier is compared with the decision tree and k nearest neighbour classifiers. Results showed that the genetic algorithm classifier outperformed both classifiers with macro-average F1 measure value equal 0.748.