Towards knowledge discovery in Semantic era

Abstract

Knowledge discovery is the non-trivial process of identifying valid, novel, potentially useful and ultimately understandable patterns in data. The complicated computational environment with ultra-large-scale, heterogeneous, highly-dynamic, and semantic-implicit data in the 21st century puts forward new problems and challenges for traditional knowledge discovery. As a solution, Semantic Web and Cloud Computing addresses the problem of semantic interoperability and large-scale resources sharing, thus making it a proper environment for future's knowledge discovery and data mining. As a probing research, this paper aims to answer the question: “what is knowledge discovery in Semantic era”. We discuss the virtual organization of knowledge discovery in Semantic Era, and introduce five roles in this environment. Moreover, we emphasize four distinguishing characteristics of knowledge discovery in Semantic era: 1) the dynamic semantic extension and self-description of algorithm; 2) the semantic integration of heterogeneous data; 3) the enablement of high-level semantic reasoning and knowledge discovery; 4) the circular refinement of knowledge and semantics. Considering the approaching era of Semantic Web and Cloud Computing, the walking towards knowledge discovery in Semantic era could be expected.