Identifying the k Best Targets for an Advertisement Campaign via Online Social Networks

Knowledge Extraction from Biological and Social Graphs

“…The first technique applied to brand-affinity matching that is not an action-based approach has been presented in [15]. In this work [19] the author proposes a general framework for the recommendation of possible customers (users) to advertisers (e.g., brands) based on the comparison between OSN profiles. This approach belongs to the first category, discussed above, to the best of our knowledge the only techniques applied to brand-affinity matching are action-based approaches.…”

“…The main goal of the proposed approach [15,19] is to identify the most suitable k possible buyers to whom distributing a given advertisement campaign. To this aim, two important aspects have to be taken into account:…”

“…where |w ij | is the frequency of the term w ij in the text T j and |T j | is the number of words in T j , and: IDF (w ij ) = log m h , where h ≤ m is the number of texts where w ij occurs. The following definition used in [15,19] is needed in order to provide the notion of match between the social profiles considered here. The TF-IDF is used to weigh the importance of the words inside the text.…”

“…As usual in both supervised and unsupervised classification contexts [19,25,36,37,44,77,80,82], the "performance" of the rank view in discovering hidden functional knowledge may be evaluated by using external criteria. Here, an external criterion relies on the existence of a ranking associated to some gold standard, obtained via information not dependent on the topology of the input network.…”

“…The chapter is divided into three sections, each corresponding to three proposed approaches in previous chapters. In particular, most results of this PhD Thesis have been already pubblished in conferences proceedings [14,15,17,19], book chapters [18] and international journals [16]. The approaches described in the previous chapters have been implemented in Apache Spark (see Chapter 2.6.2).…”