User profiling for generating bids in digital signage advertising auctions

Titelangaben

Müller, Jörg ; Krüger, Antonio:
User profiling for generating bids in digital signage advertising auctions.
In: Proceedings of the International Workshop on Ubiquitous and Decentralized User Modeling : UBIDEUM 2007. - Corfu , 2007 . - S. 46-49

Volltext

Link zum Volltext (externe URL):

Weitere URLs

Abstract

In this paper we present a simple approach to adapt advertisements on digital signage to the interests of the audience. We use auctions to sell the advertising space to the highest bidding advertisement and decision theory to determine for each advertisement how much to bid. Within this framework, we use methods from content-based recommender systems to adapt the bid to the interests of the audience. Each advertisement has a set of keywords, and the history of all advertisements a user was interested in is kept. We propose to use a naive Bayes classifier to estimate the probability that a user is interested in a certain advertisement given the keywords and user’s history. We propose interaction mechanisms to generate feedback to train the estimation functions and use the m-estimate to deal with the cold start problem.