|
|
|||||
|
|
|||||
|
|
|
|
|
|
|
Overview of the proposed course content.
| Time | Room | Activity | Speaker | Description |
|---|---|---|---|---|
| 09:00 | Arrive | |||
| 09:30 | Introduction | Nigel Shadbolt | ||
| 09:45 | Context Modeling in CommonKADS | Tim Clarke | Review of the tools available to model the focus task area, the organizational environment in which it is embedded, and the agents who perform the task. | |
| 10.45 | Coffee | |||
| 11:00 | The CommonKADS Knowledge Model | Paul Smart | The CommonKADS knowledge model details the knowledge types and structures which one encounters in a particular problem domain. The primary aim of the knowledge model is to detail the results of knowledge elicitation undertaken for a particular task area. The current session describes the basic structure of the knowledge model and the modeling formalisms that are used to represent the knowledge infrastructure of a focus task area. | |
| 12.00 | Lunch | |||
| 13.00 | Reusable Components | Tim Clarke | CommonKADS features a number of re-usable components which can serve to increase the efficiency, quality and reliability of a knowledge system solution. Re-usable components include a library of task templates, a catalogue of inference descriptions and provisional domain schema specifications. | |
| 14.00 | A Worked Example: The Sisyphus III Initiative | Paul Smart | The current session will focus on a small, but illustrative, knowledge-intensive task area with the express aim of demonstrating the practical application of the CommonKADS methodology to a real-world problem. Click here for further details. | |
| 15.00 | Coffee | |||
| 15.15 | Communication & Design Modeling | Paul Smart | The CommonKADS communication model details the communication links between agents that must cooperate in the performance of a shared task. Together, the communication and knowledge model provide input to system specification and design issues which are detailed in the design model. | |
| 16.00 | Software Tools To Support the Knowledge Engineering Process | Clive Emberey | Epistemics are a leading
provider of software to support the knowledge elicitation process. The
current session will review the advantages of software-assisted KA and the
tools available to support the knowledge acquisition process. Current software development initiatives within Epistemics also aim to deliver a new suite of knowledge editing and compilation tools which will facilitate the transition from knowledge-level descriptions to functional prototype reasoning systems. The current session will detail the steps towards this objective and present the results of early work in the specification of a meta-level model of the CommonKADS domain. |
|
| 16.45 | Finish | |||
| Evening activities: Meal at Highfield House Hotel | ||||
| Time | Room | Activity | Speaker | Description |
|---|---|---|---|---|
| 09:00 | Arrive | |||
| 09:30 | Introduction | Nick Milton | ||
| 09.45 | Problems in Acquiring Knowledge | Nick Milton | In practice, a range of problems may be encountered when trying to elicit knowledge from domain experts. This session aims to highlight common problems in knowledge acquisition, and offers some practical solutions. | |
| 10.30 | Coffee | |||
| 10.45 | Natural Techniques - Workshop | Nick Milton | A workshop providing experience in a range of 'natural' knowledge acquisition techniques, e.g. structured interview, protocol analysis. | |
| 11.30 | Contrived Techniques - Presentation | Nick Milton | In contrast to the 'natural' techniques, contrived techniques elicit information in a manner that may well be unfamiliar to the expert. These techniques include repertory grid, laddering, concept sorting. | |
| 12.30 | Lunch | |||
| 13.30 | Contrived Techniques - Practical | Nick Milton | A practical session concentrating on the application of the contrived knowledge acquisition techniques described in the previous session. | |
| 14.45 | Coffee | |||
| 15.00 | Modeling Techniques Using PC-PACK | Nick Milton | PC-PACK is an integrated tool-kit that assists the knowledge engineer in the knowledge elicitation process. The combination of a diverse range of tools within one software package enables the knowledge elicitation engineer a high degree of flexibility in the analysis and representation of knowledge structures. | |
| 16.00 | Finish |
