An Innovative Research Framework on Intelligent Text Data Classification System Using Genetic Algorithm

Recent years have witnessed an astronomical growth in the amount of textual information available both on the web and institutional wise document repositories. As a result, text mining has become extremely prevalent and processing of textual information from such repositories got the focus of the current age researchers. Indeed, in the researcher front of text analysis, there are numerous cutting edge applications are available for text mining. More specifically, the classification oriented text mining has been gaining more attention as it concentrates measures like coverage and accuracy. Along with the huge volume of data, the aspirations of the user are growing far higher than the human capacity, thus, an automated and competitive intelligent systems are essential for reliable text analysis. Towards this, the authors in the present paper propose an Intelligent Text Data Classification System (ITDCS) which is designed in the light of biological nature of genetic approach and able to acquire computational intelligence accurately. Initially, ITDCS focusses on preparing structured data from the huge volume of unstructured data with its procedural steps and filter methods. Subsequently, it emphasises on classifying the text data into labelled classes using KNN classification based on the selection of best features derived by genetic algorithm. In this process, it specially concentrates on adding the power of intelligence to the classifier using together with the biological parts namely, encoding strategy, fitness function and operators of genetic algorithm. The integration of all biological components of genetic algorithm in ITDCS significantly improves the accuracy and reduces the misclassification rate in classifying the text data.