Computational advertising in social networks: an opinion mining-based approach

In the last decade Web advertising has rapidly grown, becoming one of the major profit source for web-based businesses. One of the main challenges is the creation of effective and engaging advertisements able to attract potential customers on the product websites, leading them just one step before the product purchasing. A promising solution is the combination of user preferences from the content provided in social media and the extraction of significant aspects discussed on review websites. In this paper, we propose a 3-phase model for the content analysis from product review sites, which considers in tandem (i) the aspects discussed by users on reviews websites, (ii) the polarity of the opinions associated with such aspects, and (iii) the user profile on social networks aiming at the detection of the most "interesting" aspects and at using them for generating attractive messages. The effectiveness of the proposed approach has been validated in a real-world scenario by working with two Twitter seeds about domain-specific magazines and the data of the respective followers.