A Novel Text Clustering Approach Based on Bacterial Colony Optimization

A huge number of text documents are divided into a limited number of groups using the unsupervised learning technique known as text clustering. While the clusters contain different text documents, each cluster contains similar documents. Swarm intelligence (SI) optimization methods have been successfully used to resolve a variety of optimization issues, including difficulties with text document grouping. However, the traditional SI algorithms have many drawbacks including the local optima, low convergence rate, and low accuracy. Hence, the present research work proposes a novel text clustering approach based on bacterial colony optimization (BCO) for separating text documents based on similarity. Three separate text document datasets are used for the experiments, and performance is assessed using three different performance measures. The novel BCO approach produces high accuracy and a quick convergence rate, according to the analysis of the results. The suggested BCO text clustering strategy compares several text clustering methods to analyze strength and robustness.