Towards Cloud-Based Knowledge Capturing Based on Natural Language Processing

The organized capturing and sharing of knowledge is very important, and a lot of tools, such as wikis, social communities and knowledge-management or e-learning portals, exist for supporting this purpose. The community content- and knowledge-capturing, management and sharing portal of the European project “Realising an Applied Gaming Eco-system” (RAGE) combines such tools. The goal of the RAGE project is to boost the collaborative knowledge asset management for software development in European applied gaming (AG) research and development (R&D). To support this process, the so-called RAGE ecosystem implements a portal to support the related asset, content and knowledge exchange between diverse actors in AG communities. Therefore, the community portal in RAGE is designed as a so-called ecosystem and is intended to provide its users different tools for the capturing, management, and sharing of knowledge. In this study, we rely on the term and model definition of spiraling knowledge exchange between explicit and tacit knowledge given by Nonaka and Takeuchi.1 To achieve the goal of extracting, i.e., externalizing and explicitly representing and sharing this knowledge to its users, we propose to generate a taxonomy for faceted search automatically by extracting named entities form the knowledge sources and to classify documents using Support Vector Machines (SVM). In this paper we present our architectural approach for the NLP-based IR concepts and discuss how cloud services based on data distribution and cloud computing can improve the outcome of our system.