Abstract

Text classification is an issue for some applications. The point of text classification is to arrange text to one or many classes. Web-based media gives ample data to concentrate on individuals' thoughts and feelings about occasions in this world. The issue is to arrange texts dependent on the importance degree to the picked occasion, where the pertinence degree can be highly relevant, moderate relevant, low relevant, or irrelevant. In this paper, this issue is tackled by utilizing a classification system based on fuzzy logic and genetic algorithm. The proposed framework goes through four phases that are information assortment, preprocessing, features extraction, and a classification stage. In the information assortment stage, this framework is rely upon the Twitter text as a contextual analysis. The aftereffects of the proposed framework contrasted with and fuzzy logic-based method and Naïve Bayes classifier dependent on the adjustment rate and gradual rate. The amendment pace of proposed framework for every informational index are (98.75%, 99.45%, 99.31%, 98.70%, 98.55%) however the remedy rate fuzzy logic technique are (99.2%, 99.3%, 98.3%, 98.3%, 99.1%) and Naïve Bayes classifier are (95.7, 97.7, 98.4, 96.3, 96.7) in grouping. At the gradual rate, the proposed framework can extricate tweets more than this technique, where in dataset 1 the number of the tweets removed by the proposed framework is 160 tweets however the quantity of the tweets that separated by the Keyword search strategy, Naïve Bayes classifier and fluffy rationale based strategy are 98, 133 and 141 in a grouping.

Keywords

The article was added to IASJ on 2022-12-12

17 Total full text downloads since the date of addition