Issue No. 09 - September (2006 vol. 18)
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/TKDE.2006.144
The growing number of mobile subscribers has attracted firms to invent newer strategies to reach prospective customers in innovative but nonintrusive ways. While customer mobility creates an opportunity to reach them at desired times and locations, in practice, real-time ad deliveries are difficult due to the size of the ad-delivery decision problem. This research aims at analyzing and developing a decision policy for delivering ads on mobile devices such as cell phones. We look at the ad delivery problem from the perspective of an advertising firm, which delivers ads on behalf of its clients (merchants) to mobile customer using a wireless carrier's infrastructure. We formulate the mobile ad delivery problem as a Markov decision process (MDP) model. The ad delivery policy depends on the customer's desire and willingness to receive ads, their real-time locations, historical mobility patterns, available network capacity, and fee structure agreed upon for ad deliveries. We also develop a fast heuristic to solve larger size problems. We test our heuristic against an upper bound we developed and analyze performance using simulation.
Ad targeting, Markov decision process, wireless advertising, mobile advertising, wireless networks.
A. K. Tripathi and S. K. Nair, "Mobile Advertising in Capacitated Wireless Networks," in IEEE Transactions on Knowledge & Data Engineering, vol. 18, no. , pp. 1284-1296, 2006.