An Evolvable Rule-Based E-mail Agent

The Semantic Web is a “living organism”, which combines autonomously evolving data sources/knowledge repositories. This dynamic character of the Semantic Web requires (declarative) languages and mechanisms for specifying its maintenance and evolution. For example, for changing the behaviour of a data source, so that a new rule becomes into effect, one should not be concerned with the complex, interrelated, and dynamically obtained knowledge, and should have a way to simply specify what knowledge is to be changed. This requires the existence of a language for exacting such changes (or updates), which takes in consideration the addition/deletion and changes of rules, thereby automating the task of dealing with inconsistencies arising from those updates. To address this issue, we resort to recent developments in the field of Logic Programming, and show how the framework of EVOLP (EVOlving Logic Programs) can be put to work to model such reactive and updateable rule bases, bringing an important added value to RuleML. We make our case by exhibiting a detailed application example of how EVOLP can be used to express updateable RuleML rule bases, employing it to define an evolving e-mail Personal Assistant Agent.