Abstract: - Human life has drastically changed due to the increase in the popularity and growth of the mobile devices and mobile apps. If the number of smartphone apps continues to rise, discovering apps for people will only become more difficult. We suggest an app-based recommender that integrates the textual data of social media i.e. Sent posts as well as user interest. We apply topic modelling on social media data to derive topics, and the features of the apps. In addition, user interests are then taken into consideration when constructing the profile. All of the app's subject distributions, as well as the app preferences, are analysed to create customised lists of suggested content for each user. In order to find out whether the environment is different from the design, we use real-world data sets. This experiment has found that social media data such as user generated posts is successful for characterising users' interests, and the proposed application relies on that knowledge.Keywords— App recommendation, social media, transfer learning, collaborative filtering