Knowledge-Based Interaction Patterns for Semantic Spaces

In the past years, it has been investigated in multiple research projects how shared data spaces can be extended with concepts and technologies from the semantic web area, which led to the development of so-called semantic spaces. These semantic spaces enable clients to publish semantically annotated data at a shared repository and to react to changes of information that can be inferred from this data. The declarative description of logical dependencies and their automated evaluation through the semantic space allow for the development of simple and elegant solutions for various complex communication and coordination problems. The investigation of many space based applications as well as experience we gained during our own work have shown that there already exist several best practices for solving many of these problems. However, no standard solutions have been established yet. In this paper, we determine a set of API operations that are supported by most available semantic space implementations and employ this set of operations as a common language for the description of four generic knowledge-based interaction patterns. The paper introduces certain usage scenarios, illustrates the benefits of using semantic spaces, and should serve as a basis for further pattern descriptions.