ArsEmotica for arsmeteo.org: Emotion-Driven Exploration of Online Art Collections

In this paper, we present an application framework, ArsEmotica 2.0, where semantic technologies, linked data and natural language processing techniques are exploited for investigating the emotional aspects of cultural heritage artifacts, based on user generated contents collected in art social platforms. We rely on affective categorization models expressed by an ontology well grounded in psychology and encoded in W3C standard specification languages. We present the implementation and exploitation of the framework on a real dataset, the ArsMeteo online collection, aiming to adapt human access to cultural heritage collections by visualising emotions detected in artworks, and to let computer access by means of Linked Open Data. The use of semantic technologies enables both automatic reasoning on the elicited affective information, and interoperability or integration of tools developed within the Semantic Web and Linked Data Community.