Abstract. We are surrounded by ever growing volumes of unstructured and weakly-structured information, and for a human being, domain expert or not, it is nearly impossible to read, understand and categorize such information in a fair amount of time. Moreover, different user categories have different expectations: final users need easy-to-use tools and services for specific tasks, knowledge engineers require robust tools for knowledge acquisition, knowledge categorization and semantic resources development, while semantic applications developers demand for flexible frameworks for fast and easy, standardized development of complex applications. This work represents an experience report on the use of the CODA framework for rapid prototyping and deployment of knowledge acquisition systems for RDF. The system integrates independent NLP tools and custom libraries complying with UIMA standards. For our experiment a document set has been processed to populate the AGROVOC thesaurus with two new relationships.

The world-wide emerging e-society generates new ways to communicate among people with different cultures and backgrounds. Communication systems as forums, blogs, and comments are widely used being easily accessible to end users. Studying and interpreting user generated data/text available on the Internet is a complex and time consuming duty for any human analyst. This study proposes an interdisciplinary approach to modeling the flaming phenomenon (hot, aggressive discussions) in on-line Italian forums. The model is based on the analysis of psycho/cognitive/linguistic interaction modalities among participants to web communities and on state-of-the art machine learning techniques and natural language processing technology. This research gives the opportunity to better understand and model the dynamics of web forums, including the language involved, the interaction between users, the relation between topic and users, language intensity and differences in behavior by age and gender

No abstract