Please use this identifier to cite or link to this item: http://hdl.handle.net/10397/83303
Title: Development and visualization of linguistic query for proximity in geographical information systems
Authors: Ng, Ching-pang
Degree: M.Phil.
Issue Date: 1998
Abstract: Geographical Information System (GIS) is recognized as a powerful tool to handle spatial-related data for various geo-applications. The analytical process facilitated are commonly emphasized on statistical method, numerical model and Boolean logic. Such approaches are recently revealed having limitation in the way of problem solving and presentation of spatial information. In fact, features in real space are normally vague in shape and classification. Human interpretation of space is also multi-contextual and fuzzy in nature. It is recognized that the conventional reasoning approaches in GIS do not effectively support intuitive spatial reasoning. This thesis aims to enhance the functions of current GIS by developing a linguistic spatial reasoning approach and coupling with corresponding visualization mechanism. An approach of linguistic query is developed in this study which allows the GIS users to express the vagueness in their reasoning mind. For example, the system allows the user to search the apartment records which are 'near' to the shopping centres and 'within about 20 minutes walking distance' to the business district. Besides, it offers the interpretation of imprecision in spatial phenomenona. Proximity is the focus of spatial phenomenon in this study as a demonstration for linguistic spatial query. Fuzzy set theory is used as the theoretical basis to support the formalism of linguistic query. Different kinds of fuzzy number, as well as the logic operations 'AND' and 'OR' are discussed. The context of proximity are studied in terms of their influence in defining the linguistic query, such as topology of road network, scale, task nature of the travelling, personal urgency, etc. Qualitative and quantitative natures of proximity information are integrated by transforming to fuzzy number and facilitate the definition of linguistic criteria. Visualization of the linguistic query results is another major focus in this study. The gradation pattern generated from the linguistic query requires exploration mechanism. The Hue, Lightness, Saturation (HLS) color systems are adopted to visualize this spatial pattern generated from the query. The interesting point is that for which combination of these three dimensions is most logical to depict the gradation of imprecision. A questionnaire survey is conducted for determining the optimum combination of the three color dimensions. The linguistic query and the visualization mechanism are implemented in ArcView 3.0 with the Network Analyst module. The query of proximity is carried out in a road network. The definition of linguistic criterions by the users's preference are presented. Essentially, topology of the network is considered. The route distance rather than geometric distance is computed. By the visualized result, we can see that 'faster does not always mean shorter'. Two adjacent hue sequences and lightness change are used to depict the gradation of proximity in the road network catchment. Although the thesis presents the contextual knowledge to define proximity linguistic query, it is difficult to include all of them in the query process, especially for those related to individual experience or urgency. Besides, the vector model in the GIS package requires a long processing time to generate smooth gradation of color for depicting the query result. Nevertheless, the result shows that it is a significant achievement to enforce the current approach of spatial query by incorporating linguistic notion with the sequential color visualization mechanism. Both of the vagueness from the geographic space and human interpretation can be absorbed by the linguistic notion. The HLS color system is effective in visualizing the linguistic gradation pattern and providing the users a mental figure for the property of the region.
Subjects: Geographic information systems
Geographical perception
Space perception
Hong Kong Polytechnic University -- Dissertations
Pages: ix, 95, [3] leaves : ill., maps ; 30 cm
Appears in Collections:Thesis

Show full item record

Page views

42
Last Week
0
Last month
Citations as of Mar 24, 2024

Google ScholarTM

Check


Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.