Abstract:Text clustering has been widely utilized with the aim of partitioning specific document collection into different subsets using homogeneity/heterogeneity criteria. It has also become a very complicated area of research, including pattern recognition, information retrieval, and text mining. Metaheuristics are typically used as efficient approaches for the text clustering problem. The multi-verse optimizer algorithm (MVO) involves a stochastic population-based algorithm. It has been recently proposed and successfully utilized to tackle many hard optimization problems. However, a recently applied research trend involves hybridizing two or more algorithms with the aim of obtaining a superior solution regarding the problems of optimization. In this paper, a new hybrid of MVO algorithm with the K-means clustering algorithm is proposed, i.e., the H-MVO algorithm with the aims of enhancing the quality of initial candidate solutions, as well as the best solution, which is produced by MVO at each iteration. This hybrid algorithm aims at improving the global (diversification) ability of the search and finding a better cluster partition. The proposed H-MVO effectiveness was tested on five standard datasets, which are used in the domain of data clustering, as well as six standard text datasets, which are utilized in the domain of text document clustering, in addition to two scientific articles' datasets. The experiments showed that K-means hybridized MVO improves the results in terms of high convergence rate, accuracy, error rate, purity, entropy, recall, precision, and F-measure criteria. In general, H-MVO has outperformed or at least proven to be highly competitive compared to the original MVO algorithm and with well-known optimization algorithms like KHA, HS, PSO, GA, H-PSO, and H-GA and the clustering techniques like K-mean, K-mean++, DBSCAN, agglomerative, and spectral clustering techniques.

Multi-objective memetic differential evolution optimization algorithm for text clustering problems

Multimodal Multiobjective Differential Evolution Algorithm Based on Enhanced Decision Space Search

A novel hybrid multi-verse optimizer with K-means for text documents clustering

Adaptive Memetic Computing for Evolutionary Multiobjective Optimization

Clustering Time-Evolving Data Using an Efficient Differential Evolution.

A novel clustering-based evolutionary algorithm with objective space decomposition for multi/many-objective optimization

Multi-Objective Automatic Clustering Algorithm Based on Evolutionary Multi-Tasking Optimization

Multi-clustering Via Evolutionary Multi-Objective Optimization

Advances in Meta-Heuristic Optimization Algorithms in Big Data Text Clustering

Memetic Differential Evolution Methods for Semi-Supervised Clustering

An Evolutionary Algorithm with Clustering-based Selection Strategies for Multi-objective Optimization

Multimodal Multiobjective Evolutionary Optimization With Dual Clustering in Decision and Objective Spaces

An adaptive evolutionary multi-objective clustering based on the data properties of the base partitions

Two-Stage Evolutionary Algorithm Using Clustering for Multimodal Multi-objective Optimization with Imbalance Convergence and Diversity

A Multi-Objective Evolutionary Algorithm With Hierarchical Clustering-Based Selection

An improved differential evolution algorithm for multi-modal multi-objective optimization

Adaptive Memetic Algorithm Based Evolutionary Multi-tasking Single-Objective Optimization

A hybrid approach using genetic algorithm and the differential evolution heuristic for enhanced initialization of the k-means algorithm with applications in text clustering

Evolutionary Multi-Tasking Optimization for High-Efficiency Time Series Data Clustering

A preliminary survey on optimized multiobjective metaheuristic methods for data clustering using evolutionary approaches

A Hybrid Composite Differential Evolution and Multiobjective Particle Swarm Optimization Evolutionary Algorithm and Its Application