Prototype Implementation


We have implemented a prototype of the MIDAS middleware.
This section provides some details and insights about the current prototype implementation.

The prototype is being tested in some example scenarios to evaluate its effectiveness and efficiency.


Interaction Flows


The basic data flow between the various middleware components is shown below.
Note: For the sake of clarity we have only represented the case of a service query that is compliant with user's requirements as expressed in the profile


Data Flow Schema


Single Middleware Components


All the middleware components have been implemented in Java. Various free exisitng tools have been integrated to achieve specific functionalities.

  • The Metadata Manager (MM) is being implemented by integrating and extending the Protégé ontology editor.

  • The Context Manager (CM) is currently being re-implemented.[more details]

  • The Discovery Manager-Semantic Matching Engine (SME) has been developed by exploiting the Semantic Web Framework Jena as the ontology models repository and Pellet as the OWL reasoner.
    SME implements a matchmaking algorithm.

  • The Discovery Manager-Query Processing Manager (QPM) is implemented in two modules, which execute on a mobile node (client module) and on a fixed node (server module), respectively.

    The client module is responsible for interacting with the user by displaying services available within the user personal view, collecting her request for service and finally displaying the results of the user search. This module is implemented as a Java GUI.

    The server module is in charge of coordinating with SVM to provide the user with the list of services in the current view and with SME to process her request via semantic matching.

  • The code of the SME and QPM prototype implementation is available on request.


Last updated 16-10-2007