SMURF-BUD: Hybrid Deep Learning Based Predictive Modelling for Business Development via Social Media Analysis

Social media is becoming an increasingly useful tool for businesses to analyze and gain insight for decision-making. Social media (SM) analytics gather information from various SM platforms and analyze the SM data to understand customer preferences and market trends for making decisions in business development. Therefore, a novel Social Media analysis Using hybRid deep learning For BUsiness Decision making (SMURF-BUD) model has been proposed in this paper, which employs social media comments for business development. The proposed model uses the non-negative matrix factorization model for topic modelling, which reduces the dimensionality and enhances the accuracy in classifying customer sentiments. The proposed method uses a Dilated Convolutional Neural Network-based Bidirectional Gated Recurrent Unit (DCNN-BiGRU), which integrates DCNN for feature extraction and BiGRU for sequential learning and customer sentiments classification. The integration of these models enhances the predictive capability and improves accuracy and efficiency compared to traditional models. The proposed work has been evaluated using the customer feedback and Online Product Reviews Analysis datasets. The proposed SMURF-BUD technique has been simulated using a Python simulator for various metrics such as accuracy, precision, recall, F1-score, coherence score, and silhouette score. The proposed technique attains a higher accuracy of 99.26%, whereas previous methods such as BD-SMAB, BDMS, CIB-PA and Decision support framework achieve an accuracy of 93.54%, 93.7%, 95.16% and 97.14%, respectively. In the real world, the SMURF-BUD model helps businesses for providing market trends and understanding customer satisfaction, and develops a brand by making better decisions.