Sri Lanka Telecom

Sri Lanka Telecom (SLT) provides fixed line, TV, and broadband services to its customers in the telecommunication sector. In the fiercely competitive environment, it is important to model customers’ preferences for these services.  Using the vast amount of recorded data, we built a recommendation system which allows SLT to improve its cross-selling and up-selling to increase revenue. A final-year undergraduate at the Department of Industrial Management (DIM) was the main contributor to the project as part of his final-year research. DIM and SLT had knowledge-sharing sessions on a weekly basis to enhance the outcomes of the research. Once the recommender system was built and tested, DIM passed it to SLT to integrate into their product. This proposed recommender system was published as a research paper.

Hewawitharna, Chiran & Rajapakse, Chathura & Asanka, PPG. (2021). Wide and Deep Learning for Enhancing Context-Aware Recommender Systems in the Telecommunication Industry. : https://www.researchgate.net/publication/372400593_Wide_and_Deep_Learning_for_Enhancing_Context-Aware_Recommender_Systems_in_the_Telecommunication_Industry/citations

The telecommunication sector faces a significant challenge due to defaulting customers who fail to pay their dues on time. Sri Lanka Telecom (SLT) consulted the Department of Industrial Management (DIM) at the University of Kelaniya to help them identify the defaulters that enable them to take proactive measures to mitigate the risk. DIM explored a feature list for identifying defaulters and developed multiple predictive models for fixed-line telecommunication as a part of undergraduate research. DIM handed over the outperforming default prediction model for implementation to SLT, enabling SLT to mitigate the risk associated with defaulters. DIM disseminated the default predictive model through a Scopus-indexed publication:

 S. Ginige, C. Rajapakse, D. Asanka and T. Mahanama, "Defaulter Prediction in the Fixed-line Telecommunication Sector using Machine Learning," 2023 International Research Conference on Smart Computing and Systems Engineering (SCSE), Kelaniya, Sri Lanka, 2023, pp. 1-8, doi: 10.1109/SCSE59836.2023.10214995.