A recommendation system for offers in telecommunications
Cong Dan Pham, Tuan Anh Chu, Huy Hung Pham, Manh Linh Dao, Thanh Son Pham, Van Hung Trinh, Duc Hai Nguyen · 2021
Recommender Systems are recently applied in many various areas such as entertainment, education, science, and especially e-commerce. In this paper, we propose an approach for building a recommender system in telecommunications for the telecom data about the historical behavior of consumers in six months. Our collected data spans over 6 months and contains how many times each offer has been purchased by a customer in each month and revenue generated by him/her in that month. Our recommender system consists of three main modules. In the segmentation module, we segment our customers into homogeneous groups, consumption wise. In the recommendation module, we utilize the Alternative Least Squares method to determine candidate offers. In the business rule module, we apply business rules to finalize the best telecom offers out of these candidate offers to recommend to each customer. Thanks to big data platforms, our recommender system attains a running time of approximately 10 seconds to 15 seconds to handle one million customers.