Abstract
Abandon Rate is an effective evaluation indicator for telecommunication operators to measure network performance. It is the percentage of incoming calls made to a call or service center that are dropped by the callers before they are being attended to by a call center representative (agent). The research focuses mainly on improving abandon rate forecasting problems and offers an intelligent hybrid model in which decision rules will be generated as programs that will be able to reason and use a knowledge base to classify and forecast abandon rate. This proposed hybrid model is constituted by five artificial intelligence (AI) tools: experiential knowledge (EK), Discretization method - Minimize Entropy Principle Approach (MEPA), fuzzy set theory (FST), rule filter (RF), and rough set theory (RST). The model will be validated using obtained data from Airtel call center collected over a three months period. The proposed model is expected to improve forecasting performance in terms of higher accuracy, using fewer attribute sets in rule generation, and generates fewer numbers of decision rules. Keywords: Abandon rate, Classifier, Experiential knowledge, LEM2 algorithm, MEPA.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
