Multidimensional Author Profiling for Social Business Intelligence

Abstract This paper presents a novel author profiling method specially aimed at classifying social network users into the multidimensional perspectives for business intelligence (SBI) applications. In this scenario, being user profiles defined on demand each particular SBI application, we cannot assume existence of labelled datasets training purposes. Thus, propose an unsupervised to obtain required profile classifiers. Contrary other approaches in literature, only make use users’ descriptions, which are usually part metadata posts. We exhaustively evaluated proposed under four different tasks along with state-of-the-art text achieved performances around 88% and 98% F1 score gold standard silver respectively. Additionally, compare our results supervised previously two tasks, getting very close despite using method. To best knowledge, is first designed label way classifiers similar performance fully ones.