Towards CSpaces: A New Perspective for the Semantic Web

Information overload is mainly the result of the combination of four factors: the enormous amount of information available; the heterogeneity of information sources and information channels; the generation of a significant percentage of redundant information; and inefficient mechanisms for filtering, searching and classifying information. Given that the former factor cannot be changed, and current forecast expects that information grows exponentially in the next years, research and industry efforts are focusing to overcome the other three. The association of machine-understandable semantics to formally describe data published on the Web and the development of appropriate tools that can handle this method to describe data are the approaches that the promoters of the Semantic Web have suggested to overcome the problem of information overload in the Web. Although, the Semantic Web promises a new level of service with regard to the current Web, a more drastic approach is required. Conceptual Spaces (CSpaces) envision the future of the Semantic Web as a cooperative environment where communication between humans, machines, and human-and-machines will be reduced to the acts of publishing and reading machine processable semantics in a persistent collection of individual and shared information spaces. Decreasing the amount of syntactic data representation in the Semantic Web, and therefore, make machine processable semantics the prevalent representation formalism will facilitate interoperation between heterogeneous applications, web services, agents, humans and so on. Natural language generation and graphical knowledge visualization techniques will make possible that humans deal with this "purest semantic" Web. In addition, CSpaces will also decrease redundancy of the information stored and will provide a better organization of the data articulated around ontologies.