Resumen de Cam-odl: a semantic a semantic collaborative awareness model for web agents
Manuel Salvadores Olaizola
The concept of Virtual Organisations (VOs) has rapidly grown as an important subject of study and it has been applied to different areas of computing. Due to its abstraction mechanisms to create, maintain, resolve and dissolve; VOs have emerged as a key aspect on Collaboration and Cooperation Systems. This thesis models the creation phase of VOs and provides a consistent framework based on the Spatial Model of Interaction and Se- mantic Web technologies. The Semantic Web is an extension of the Web to enhance its capability of sharing data. Through RDF we can provide an unique representation of the data in a structured format. In the current syntactic Web URIs are used just to fetch documents, or parts of them. The Semantic Web uses URIs to name things (subjects) and their relationships (predicates) to other things (objects), in that sense RDF is a list of (subject,predicate,object) structures. Based on the Semantic Web and RDF the Open Linked Data (OLD) has come into view as a multi-domain decentralized knowledge base. The OLD is one of the tangible products of the Semantic Web and is reaching end-users by enabling applications to browse its interconnected data sets.
Current algorithms for constructing VOs focus on the development of complex heuristics to determine the best agent coalitions. Our research makes new grounds towards enhancing these algorithms by means of a Semantic Web framework. This framework called CAM- ODL is based on the Collaborative Awareness Model (CAM). CAM is based on previous Computer Supported Cooperative Work (CSCW) research. CSCW are characterized by their ability to support and manage large numbers of coordinated heterogeneous resources and services while they cooperate to accomplish a common goal. CAM has been designed, from the beginning to be a parametrical, generic, open, model that could be extended and easily adapted. This model allows managing not just resources and information but also interaction and awareness. More specifically, CAM allows: i) controlling the user interaction; ii) guiding the awareness towards specific users and resources; iii) scaling interaction through the awareness concept. CAM-ODL enhances CAM by embracing the Semantic Web and its paradigms providing inter-operability with the Web of Data.
CAM-ODL as the outcome of this research is built upon the following components: (a) an ontology, (b) a set of inference rules and (c) an interface to standardize the interaction between agents. These three components make use of Semantic Web technologies providing a consistent framework that leads the construction of VOs. The model developed is validated through the course of this thesis both theoretically and in real use case scenarios. These use cases are based on heterogeneous scenarios that prove the flexibility of the model and its ability to construct VOs. In this thesis are introduced two different scenarios: the first scenario is a Semantic Grid cooperation where CAM-ODL plays an important role in order to create a suitable Grid of agents that can resolve a specific task, the second scenario introduces an interaction with the OLD, this scenario uses the Friend of a Friend (FOAF) profiles in order to create VOs of people that know each other and can cooperate to resolve a problem.
CAM-ODL brings together CAM and the Semantic Web enhancing the process of VO construction. The solution presented in this thesis opens a wide range of possibilities since it is fundamental for Web Agents to have a perception of its surroundings, and CAM-ODL improves their perception by making them inter-operable with other models and with the Web of Data.
