Tourism Destination Recommendation System on Social Media X (Twitter) with Content-Based Filtering (CBF) and Gated Recurrent Unit (GRU) Approach

Along with the growth of the tourism sector in the city of Bandung, challenges have emerged within the community. People often struggle to find tourist destinations that align with their preferences. One way the community seeks information about tourist destinations is by utilizing social media platform X (Twitter). However, information about a tourist destination obtained from Twitter is considered less effective in helping people find suitable tourist destinations. Therefore, a tourist destination recommendation system is needed to provide recommendations that match users preferences. This research aims to combining the Content-Based Filtering (CBF) method with Gated Recurrent Unit (GRU) optimized using the Root Mean Squared Propagation (RMSProp) algorithm to develop a tourist destination recommendation system in the city of Bandung based on datasets from Twitter containing a total of 99 tourist destinations and 95 users. Various types of Synthetic Minority Oversampling Technique (SMOTE) method will also be applied to overcome the problem of class imbalance in the dataset. The research results show that the combination of CBF and GRU optimized with RMSProp, along with the application of SMOTE+ENN, achieved the highest results with an accuracy of 91.44%, precision of 82.36%, recall of 90.75%, and F1-Score of 86.35%.