Influencing Consumer Behaviour on E-Commerce Platforms Through Social Media Marketing Using BiLSTM-KNN Models

E-commerce systems depend significantly on comprehending and forecasting consumer behaviour, frequently shaped by social media marketing. Current methodologies in this domain exhibit deficiencies in precision and computational efficacy. This study introduces an innovative hybrid model that integrates BiLSTM and KNN to improve customer behaviour prediction by utilizing social media sentiment and e-commerce behavioural data. The model markedly surpasses current technologies, with an accuracy of 96.5%, an F1 score of 96.0%, and a recall of 97.0%. The suggested approach adeptly incorporates sequential dependencies and neighbourhood-based categorization, rendering it a formidable instrument for forecasting consumer behaviour. The model improves performance indicators and decreases training time, rendering it extremely appropriate for practical applications in e-commerce systems.