Software agents for projects identification - The case of CDF Projects

Abstract

This paper describes and reports on a prototypical implementation of a Multi-agent system for Constituency Development Fund (CDF) Projects Identification. Agents as representatives of various I stakeholders have the task of availing the projects requirements/preferences and offers and performing the necessary updates of the system data. The approach proposed uses the semantic web languages and tools; Rule Markup Language (RuleML) to specify and publish projects requirements and preferences as per the CDF Act 2003, as the basis for agent knowledge base and Resource Description Framework (RDF) to represent the projects offerings. The system also contains a persistent storage of the Constituencies data and the yearly CDF Allocations in a MySQL Database. The use of semantic language is motivated by the need to increase openness and interoperability among agents.