Participants were asked to identify the areas of their community that they most liked and disliked by clicking on the appropriate square of the grid. The only visual assistance on the map was the name of streets. They were to use a GREEN pointer to indicate liked areas and a RED pointer to indicate disliked areas. By clicking on one of two radio buttons located on the side of the map, they could “load” the pointer with either colour. When participants clicked on the square, a small window with a question mark appeared, asking them to state their reasons for liking or disliking that area of the community. When finished, the participant clicked on a button labelled “Submit” and their input was transferred to the UIC server. Each of the participants’ selections could then be stored in the Web access logs for analysis and feedback into the planning process. Since the server was linked to an Oracle database and a GIS application, we had the capability of taking all the points that were selected by the participants, sorting them by longitude and latitude, and plotting them on a map automatically. The Oracle database could also group the associated comments. In this manner, a community-input database was created that contained the range of views about areas liked and disliked with the associated reasons. We then created a number of GIS choroplethic maps to illustrate the intensity of likes and dislikes (urban likability and urban dislikability). We used dots to represent intensity: the number of dots in each cell of the map was proportionate to the number of times that area was selected by the residents in the survey exercise. In addition, these GIS maps were interactive; clicking on an area (or cell) of the map opened a window of text that listed the residents’ stated reasons for liking or disliking that area. Since the maps provided written evaluations for each point, they were extremely useful in supplying specific directions for improvement and could easily be incorporated into the next stages of the community planning process. In a later prototype we used nested maps. We also used a grid but it advanced the initial project by dealing with a larger geographic scale. Not only does the user navigate using the grid, but he or she also selects a square of the grid to comment upon. As mentioned above, the primary disadvantage of this method is that users cannot select the specific buildings or combination of buildings that they wish to comment upon, since the size of the final selection square is predefined. The advantages are clearly evident in the above example where the uniform selection areas allowed researchers to easily create sort and analyse the users’ feedback. Figure 3, below, shows an example of the nested map method.  In evaluating the grid as a selection mechanism, it has clear advantages and disadvantages. As in the above example, the grid enables very fast analysis and compilation of spatial data that can be easily compared among participants. On the other hand, we found that users did not have enough discretion in selecting the particular areas of the community and buildings that they wished to comment on. In addition, they were constrained to the square shape of the grid: even if they only wanted to comment on one building in a corner of this large square, they had to select the entire square. As a way to compensate the shortcoming of the grid method, we introduced the freehand sketch method. This project introduces an entirely new technology for enabling user feedback using Web-based maps. In this new prototype, participants can go to the online survey Web site and use a drawing tool to select the areas of the community that they wish to comment upon; their locational choices are not limited by the pre-defined geographic areas of the square grid. On the initial screen, the participants view a structure base map of the community along with two buttons labelled “Click to select area with drawing tool” and “Click to type in comments.” When a participant clicks on the drawing button, the cursor turns into an arrow and upon pressing and holding down the left button of the mouse, it starts to draw. The user may draw any shape on the map and when the mouse is released, the lines close on themselves to form a polygon. If the user does not draw an enclosed shape, the program approximates the line that closes the shape into a polygon. Figure 4, below, shows an example of sketching capabilities on aerial photograph.  An example of sketching capabilities on aerial photograph Each of the discussed prototypes so far provides particularly features; some are designed to sketch, others to type in comments. Some are grid based and others are based on freeform drawing. Some have zooming others do not. We are in the process of developing a Web site that will combine the best design features of the interfaces described above. Users will have options to work with or without the grid, to zoom in and out, to add layers, to type in comments, to sketch, and so on. This has apparent advantages (more choices) and disadvantages (confusion, technical difficulties to create the interface and to work out the database). Figure 5, below, shows an attempt to create a “complete” software interface for two-way communication. |