The impact of social media on people’s lives has significantly grown over the last decade. Individuals use it to promote discussions and a way of acquiring data. Industries use social media to market their goods and facilities, advise and inform clients about future offers, and follow up on their direct market. It also offers vital information concerning the general emotions and sentiments directly connected to welfare and security. In this work, an improved model called Improved Sentiment Urgency Emotion Detection (ISUED) has been created based on previous work for opinion and social media mining implemented with Multinomial Naive Bayes algorithm and based on three classifiers which are sentiment analysis, urgency detection, and emotion classification. The model will be trained to improve its accuracy and F1 score so that the precision and percentage of correctly predicted texts is elevated. This model will be applied on the same data set of previous work acquired from a general business Twitter account of one of the largest chains of supermarkets in the United Kingdom to be able to see what sentiments and emotions can be detected and how urgent they are.