The COVID-19 epidemic has changed consumer attitudes, beliefs, and emotions, therefore influencing either increasing or decreasing mortality and severity. Unexpected pandemic effects have changed public health dynamics, consumer behaviour, and purchasing patterns, as well as public health policies. This paper explores these developments using advanced machine learning and sentiment analysis. This paper shows how the epidemic has altered consumer feelings and choices using sentiment analysis, more notably, “lifestyles”. This approach automatically generates a prediction model logic and detects latent trends in giant data by means of machine learning. Quickly acquired and analysed using mining methods, the data provides immediate understanding. The study uses semantic orientation to classify opinions as either neutral, positive, or negative in order to understand consumer sentiment. This classification shows consumers’ perspectives on the epidemic—that of optimism, worry, or ambivalence. The study also applied K-means clustering—a sophisticated unsupervised machine learning technique—to order consumer sentiment data. Through an appropriate centroid count, the researchers can identify patterns and categorise consumers based on attitude. This clustering method allows for more precise analysis of pandemic-related emotional reactions and lifestyle changes. The output of this work emphasises the application of machine learning for sentiment analysis. The project intends to help businesses, politicians, and academics navigate the post-pandemic climate by recognising and evaluating these shifts in consumer behaviour.