Heteronomy and Ontology Co-Evolution in Information Systems

AUTHOR
Francesca Arcelli Fontana, Ferrante Formato and Remo Pareschi

ABSTRACT

The Web is the most important information framework of our generation and probably the next and the Semantic Web is a project to make the Web fully “understandable”. At the current stage it stays very much a “project” shared by academic institutions and research centers. On the other hand, the primeval Web (so called Web 1.0) has evolved on its own into something completely different, Web 2.0. Without getting into a protracted argument over the exact definition of “Web 2.0”, there is a general consensus that it is all about people: it doesn’t care only about technology or standards, just make sure that it is people-oriented by letting and enabling people create, collaborate, share and interact. More than that, and curiously enough, the Web 2.0 has boosted a thoroughly unexpected concept muted by biology and complex systems : co-evolution. In fact, while the nitty gritty of Semantic Web -ontologies and RDFs- still fight their way through the niche of academic and educational institutions, the hottest sites of the Network are no longer mere content containers, but lively Web communities that live their metamorphosis within global social networks like MySpace.

Complex networks are everywhere, from molecular aggregation of amynoacids to telephonic networks and motorways. It seems that complex networks are necessary characteristics of life itself, in the sense that without a complex network no form of life can be developed. One of the most bewildering things is that complex networks are never alone. For any complex network, there is another complex network somewhere that creates it, reflects it and co-evolves with it. For example the thought is a (complex) networks of concepts reflected into a (complex) aggregation of communities of neurons that reflects upon a (complex) community of people. But the most amusing thing is that the converse is difficult to deny. Therefore, any explicit specification of knowledge -according to Tom Gruber’s definition of ontology- to make sense must reflect an explicit distribution of minds that share such knowledge. As a consequence, the Web graph is the reflection of a social network which in turn is the reflection of a complex cognitive map. Thus, although the quest for semantic clarity remains high, the context and the scope of a “Semantic Web” has dramatically changed. Knowledge is no longer poured from heaven into ontologies but is negotiated through Web communities. In fact, our assumption here is that the relationship between conceptual networks (ontologies) and people networks (Web communities) is reciprocal and dynamic: ontologies identify communities and communities, through their practices and interests, define ontologies and determine their evolution. By applying proven constraints techniques for querying conceptual languages in the style of KLONE and coupling them with community detection algorithms from the tradition of complex networks, in this paper we want to show, given an ontology, how to extract a set of Web communities by matching each part of the ontology with the kernel of Web community. Furthermore, we will propose an extension of the layer cake road map for the Semantic Web, that keeps into account the relationship between people and content underlying the evolutionary step from Web 1.0 to Web 2.0.