Personalised Hyperbolic View

Thus, for the BT case study, the hyperbolic view has been modified (to a more click-and-get-the-results-interface) according to these objectives:

• The first target was to further simplify the use for the user (the user should be able to directly see the concepts and attributes he is interested in, so the most important attributes are now as well visualised as nodes), and other relations than the "is-a" hierarchy between the nodes should be made possible.

• The second target was to allow personalised views for different user groups as it is not useful to show all concepts (and attributes) to all types of users. So I have made a view for a project manager and one for a programme manager. These two user groups see only the concepts which are interesting for their daily work. As the programme manager is not interested in costs, he is not offered those nodes in his view. Furthermore, he is presented a lot less attributes under each concept.

• The third point was to enable a not exact match.

• The fourth point was to prevent the showing of internal identifiers by the programme, which are insignificant to users.

The disadvantage of this view is that it is not as flexible as the original version. Our opinion, however, is that for most users it is flexible enough and definitely more plausible to use. The proficient users who need more flexibility can still use the original version.

How the personalised Hyperbolic View of a programme manager works is explained in the following two examples: The general proceeding is that the user either chooses one or two nodes he is interested in. Then the applet creates the corresponding query. The programme logic in this is that the query contains the relation path between the first node and the centre node and, if applicable, the second node. This query is sent to Ontobroker and the results are displayed in a table.

Example 1: (one node)

The user wants to view projects of a certain theme. The user clicks on the node "themes". He chooses a theme from the list or types in part of the theme. Then the user starts the query by pressing on the corresponding button. As a result, the theme and all projects in this theme are shown. The programme has thus built a query which combines all nodes on the path between the node "themes" and the centre node ("project"). The programme knows that "themes" is an attribute of project, therefore it treats the arrow between the "themes" node and the "project" node as an attribute. (Query: FORALL p,t <- p:Project[themes->>t].)

Example 2: (two nodes)

The programme manager wants to see the description of the deliverable of a certain project. He clicks on the node "Project", types the number of the project (124PJ) into the corresponding attribute field, and clicks on the node "description" which has an arrow to the deliverable node [Figure 22]. Then he starts the query. The programme creates a query by following the path from the start node (first node) to the end node (second node). It knows the significance of each arrow connecting the nodes (either a connection to an attribute, an "is-a" hierarchy between two concepts or another relation between two concepts). First, the project is specified (p:Project[number->>"124PJ"]), then the connection between the concept "Project" and the concept "Deliverable" is realised (p:Project [has_deliverable->>d]), and then it is specified that the attribute description of the concept "Deliverable" should be displayed (d:Deliverable[description->>DESCR]). These query terms are connected by AND.

Figure 22 - Personalised Hyperbolic View for the Programme Manager

The result is seen in the table (Figure 23). If the user had not selected "exact match", the first query term would look different (p:Project[number->>n) and the query would contain the following additional term ( AND regexp("124[P/p][J/j]",n,x1)).





Broadband Satellite System Models

The Cadance simulation suite will be used to develop system models (constellation, radio channel and basic network layer) of the following systems: Teledesic, Celestri, Skybndge, Astrolink and Spaceway. Refinements and specific parameter requirements will be developed with CFD input.


Report on Broadband Satellite Simulations

Report summarising Broadband system capabilities and constraints, based on simulations run on model developed in Deliverable 1.


Upgraded Cluttered Site Model in GRASPS

Existing cluttered site model exists in GRASP7 and needs to be upgraded to GRASP8 in order to incorporate the multiple reflected ray capability. Upgrade requirements will be based on recommendations generated by the 1997/98 work. Model will also include database of

0 0

Post a comment