Introduction NeWGis is the acronym for Neural Web Gis- a route of work which forms part of our PhD researches. We are investigating the development of models for automated plan designing that learn by users interacting through the web. We are looking at the potential for integration of such models with drafting, visualisation and database systems like CAD-GIS. The motivation is two fold. First, to use the power of distributed knowledge in order to automatically generate innovative plan designs by effectively using partial and ill- defined information. Distributed knowledge is used as a mechanism to search the decision space for creative or innovative solutions. It can be seen as an extended case based repository that could be used to reveal unforeseen associations among the decision variables. The second motive relates to issues of co-operation and coordination in multi-participatory and multidisciplinary design and planning. Automatic plan design in the context of Design and Planning Support Systems can be seen as a tool to support the generation of solutions from the early phases of design and allow continuous evaluation of those solutions by the different participants through the various decision making phases. The method for artificial plan designing is presented elsewhere (www.casa.ucl.ac.uk/NeWGis/APD) and some other issues of knowledge representation are briefly discussed on www.casa.ucl.ac.uk/NeWGis/Know. Here we briefly present the overall structure and philosophy of this tool. The core of the system is based on a Web-GIS tool with added functionality for generation of plans through learning. The integration of such models with GIS functionality for storing, browsing, querying, analysing and editing spatial and descriptive information can offer greater insight to decision making processes and help us move towards the development of a more comprehensive and more interactive Design and Planning Support System. A generic architecture:
A UML sequence diagram describing the communication among the different parts:
|