Development of the Recommended Coffee Shops Application based Twitter Sentiment Analysis

Nowadays, the coffee industry trend is rapidly growing due to the demand of the consumers. It directly affects the increase in the number of coffee shops and coffee cafes. Then, the coffee shops have become the appointment place or the informal meeting area. Moreover, many people visit coffee shops and upload their opinion and feedback on social networks such as Twitter and Facebook. Feedback on social networks can be considered to be one of the decision tools to visit coffee shops. However, the limitation of the data from social networks is unstructured and uncategorized information. Therefore, our paper proposed the application to recommend a coffee shop based on Twitter data. Chiang Mai province is the focused area. This application can visualize the Twitter data on the map with all coffee shops mentioned on Twitter. In addition, the application used sentiment analysis with the help of the Naïve Bayes method. After implementing the sentiment analysis, the application can classify whether it should be recommended or not. The proposed method can perform with high performances in terms of the average accuracy, precision, recall, and F1-score.