The issue of feature subset selection for data of high dimensionality in classifying opinions has been investigated in this paper. Feature selection schemes select the crucial feature subset producing similar or superior classification outcomes compared to the original feature set obtained. Despite their higher efficiencies, wrapper-based feature selection schemes have high overheads of computation. However, the issue is the problem is Non-deterministic Polynomial (NP) hard. This work suggests the clustering scheme based on MST optimized by the Group Search Optimization (GSO) for efficient selection of features. The Amazon camera review dataset is used for evaluating the suggested scheme. Experiments are conducted to evaluate the proposed method and compared with the MRMR feature selection, FCM clustering, and MST based clustering.