Bio inspired Boolean artificial bee colony based feature selection algorithm for sentiment classification

Over the past few years, sentiment analysis has become most predominant aspect to be extracted in the modern digital era. Due to digitization all over the world, the number of existing digital documents increasing exponentially day by day. Analyzing such documents to extract the exact meaning itself becomes a tedious task. And analyzing those documents in sentimental point of view has increased the modality of difficultness to another level. In this research work, an effective bio-inspired filter-based feature selection algorithm is proposed to effectively handle the documents from sentiment point of view. A popular bio inspired algorithm called Artificial Bee Colony algorithm is modified using Boolean operators to effectively handle feature selection problem in classifying the documents from sentimental aspect. A total of 9 different documents instances are taken for evaluating the performance of proposed model. A total of 5 different filter-based measures are used to quantify the working model of the algorithm. The outputs are analyzed to state-of-the-art algorithms, and the comparative results demonstrate a substantial improvement over the existing models. © 2024 The Authors