Knowledge Extraction from Textual Resources through Semantic Web Tools and Advanced Machine Learning Algorithms for Applications in Various Domains

Nowadays there is a tremendous amount of unstructured data, often represented by texts, which is created and stored in variety of forms in many domains such as patients' health records, social networks comments, scientific publications, and so on. This volume of data represents an invaluable source of knowledge, but unfortunately it is challenging its mining for machines. At the same time, novel tools as well as advanced methodologies have been introduced in several domains, improving the efficacy and the efficiency of data-based services. Following this trend, this thesis shows how to parse data from text with Semantic Web based tools, feed data into Machine Learning methodologies, and produce services or resources to facilitate the execution of some tasks. More precisely, the use of Semantic Web technologies powered by Machine Learning algorithms has been investigated in the Healthcare and E-Learning domains through not yet experimented methodologies. Furthermore, this thesis investigates the use of some state-of-the-art tools to move data from texts to graphs for representing the knowledge contained in scientific literature. Finally, the use of a Semantic Web ontology and novel heuristics to detect insights from biological data in form of graph are presented. The thesis contributes to the scientific literature in terms of results and resources. Most of the material presented in this thesis derives from research papers published in international journals or conference proceedings.