Plateforme d'étiquetage dynamique pour la fouille d'opinion

Social networks mirror public opinion. Thus, they are of great interest for opinion mining and sentiment analysis. In most cases, sentiment is classified according to a polarity criterion or a linear gradation. Is it possible to learn more complex patterns with limited itemsets? In this article, we investigate brand equity. The goal is to analyze how social groups perceive a given brand. Brand identity is described with a structured set of adjectives, considered as semantic features. Since supervised pattern recognition requires labeled datasets, a platform was set up for social networks sentiment analysis. The platform relies on collaborative and dynamic principles designed to relevance-based filtering and semantic features reconfiguration. MOTS-CLÉS : apprentissage statistique, image de marque, réseaux sociaux, annotation, IHM, analyse de sentiments.