Sentiment Oriented Contextual Advertising

摘要

Web advertising (Online advertising), a form of advertising that usesthe World Wide Web to attract customers, has become one of the world’s mostimportant marketing channels. This paper addresses the mechanism of ContentOriented advertising (Contextual advertising), which refers to the assignment ofrelevant ads within the content of a generic web page, e.g. blogs. As blogsbecome a platform for expressing personal opinion, they naturally containvarious kinds of expressions, including both facts and comments of both apositive and negative nature. In this paper, we propose the utilization ofsentiment detection to improve Web-based contextual advertising. Theproposed SOCA (Sentiment-Oriented Contextual Advertising) framework aimsto combine contextual advertising matching with sentiment analysis to selectads that are related to the positive (and neutral) aspects of a blog and rank themaccording to their relevance. We experimentally validate our approach using aset of data that includes both real ads and actual blog pages. The results clearlyindicate that our proposed method can effectively identify those ads that arepositively correlated with the given blog pages.