Deliverable 3.3

Customers of the companies, which provides maintenance and technical support services, send fault messages in a SOAP/XML format to the maintenance center, where message data is analyzed by experts and may or may not be stored in a file system. Fault messages contain data which is potentially useful for fault analysis and predictive maintenance. To make it easily and flexibly processable, the maintenance center has to install a storage system of a new generation, which is easy to maintain, extend and query.

Based on the problem description, we have selected a semantic web-based approach to storage and retrieval of log data. Having a lot of customers and providing a large variety of services, maintenance service provider will benefit from an integral (pseudo-)centralized storage of its business data including but not limited to maintenance activities. We have selected a Semantic Web technology for implementation because it has a set of distinctive features which enforce data representation. Here we can name at least graph-based data representation, simple, but effective class-subclass and property-subproperty relationships, bringing add values, which are hard to implement using RDBMS. These features are naturally supported by most of the semantic storages and may be extended using third-party inference engines.