This thesis is a part of the knowledge module in the Ad-Hoc InfoWare project at the University of Oslo. The goal of the project is to enable information sharing between rescue personnel at an emergency scenario by the use of shared ontologies. During a rescueoperation, there are different organizations involved like police, ambulance and fire fighters. To ensure an efficient operation they need to co-operate across organizational boundaries.This creates the need for information sharing between different systems with different interfaces in a Mobile Ad-hoc Network.
Topic Maps can represent ontologies and can be realised as a knowledge layer on top of different system in different domains. The Topic Map layer will serve as a common ground for all the systems and enable the different laptops, PDAs and mobile devices to share information. The Topic Maps technology has many functions that can contribute to thedesired result. The merging mechanism is one of the cornerstones in the Topic Map standard. This mechanism removes redundant information items when two arbitrary topic maps merge. This can contribute to a global view of the different domains and tie them together across thedifferent systems.
So far, there has hardly existed any application of Topic Maps on mobile devices such as cellular phones. The Mobile Topic Viewer (MTV) is the only engine for mobile devices. There is one important aspect missing in the MTV, the process of merging topic maps. To test merging of topic maps on mobile devices, the concept of merging has been implemented in the MTV. Since the Topic Maps standard is targeting usage on resourceful devices, like laptops and PC’s, and not on mobile devices, I performed benchmark tests on the latter with different sized topic maps. A very faithful implementation of the merging algorithm can be too resource demanding for a mobile device. Therefore, the challenge was, in addition to implement the algorithm, to optimise with respect to system time and device reasonableshortcuts in the algorithm.