Abordagem multiagente para a propaganda pervasiva.

Abstract

The Pervasive Advertising stands out from other forms of ad serving by allowing the delivery of context-sensitive advertisements to consumers. Thus, it is possible to deliver the most relevant ads, increasing the effectiveness of advertising. However, delivering the most relevant ads, which have been the focus of research on Pervasive Advertising, implies in the reduction of the total delivered ads, and this reduction has a direct impact on the objectives of the advertising campaign. That is, in this search for more relevant ads to consumers, the goals of advertisers have been neglected. Thus, it is necessary to find a balance between the interests of consumers and objectives of advertisers. The assumed hypothesis in this work is that by means of a negotiation scenario, software agents, representing the consumers and the advertisers will be able to achieve the desired balance. To enable this negotiation scenario, we developed: a model and protocol of negotiation, in which advertisers and consumers establish rounds of proposals and counterproposals to each defending their own interests, obtain a balanced relationship; a model for obtaining contextual information, which allows agents to acquire in a dynamic and transparent way, the information needed to negotiate; and a model for representing the contextual information, which provides a common language for communication among agents. In order to validate the developed models, we performed a series of experiments focusing on the obtaining and using contextual information as a predominant factor for determining the relevance of the ad. The analysis of the results shows that the negotiation model is effective in finding the desired balance.