Linking entities in semantic web environment

Recently, there has been an increasing interest on entity linking which transforms text data into a higher level semantic. State-of-the-art entity linking systems rely on comprehensive knowledge bases (KB) constructed in a long period of time. But, most business domains do not have such KBs, and therefore entity linking requires a lot of effort to develop domain knowledge in those domains. State-of-the-art approach requires firstly putting that effort for developing comprehensive domain knowledge before developing an entity linking system. Contrarily, to create business focused results rapidly, we propose a new approach for management of entity linking knowledge. Our approach is an evolutionary software development methodology that depends on developing the entity linking application and the knowledge required for entity linking together. For this purpose, we adapt entity linking architecture to allow rapid prototyping and iterative development. On the other hand, Semantic Web researchers have focused on knowledge graphs in recent years and it is accepted that creating enterprise knowledge graphs presents many opportunities for business domains. In line with this purpose, entity linking helps to include text data into knowledge graphs. But, business domains without structured data require both developing KB and linking text data to the KB. In this direction, knowledge graphs can be created in a test driven and iterative way using the methodology proposed in this thesis. The thesis presents a case study that evaluates the applicability of the proposed architecture and methodology.