Applications' Startups: Predicting Future of Smartphones Applications

Start-ups are fast growing entrepreneurship. They have emerged for fast propagating of new technology, products or services. In this work, a measurement study of mobile phone applications will be conducted. This study aims to reveal the popular applications' fields in smartphones world. This popularity is utilized to recommend a field for a new startup. To this end, a new model based on supervised machine learning back propagation neural network (BPNN) model is proposed. The new model is trained and utilized to predict the popularity of new applications in any application field. To facilitate our process, a web-crawler has been written to harvest smartphones' applications information from a popular Android platform's market. The crawled data consists of download count, applications categories and users' reviews. These data has been fed to train the proposed model. Our results show that the model has been trained with a mean square error root approximately 0.002, which show the relation between download count and application category. This relation reflects the popularity of some categories in smartphone applications' market.