Semantic Data Model for Multi Thematic Spatial Databases.

Date of Submission

December 1988

Date of Award

Winter 12-12-1989

Institute Name (Publisher)

Indian Statistical Institute

Document Type

Master's Dissertation

Degree Name

Master of Technology

Subject Name

Computer Science

Department

Theoretical Statistics and Mathematics Unit (TSMU-Kolkata)

Supervisor

Gupta, Amarnath (TSMU-Kolkata; ISI)

Abstract (Summary of the Work)

Several models (1,2,3,4,5] for spatial information management have been proposed. Most of these systems offer a definition and query language which déclares a spatial object in terms of a set of geometric base objects, mainly point, line and polygon. Some of them such as GeoSAL |2] offer more complex objects such as chain of lines and tessellation of polygons. Geometric declarations have some clear advantages.• The spatial objects are directly representable in geometric data structures.• Definite spatial access mechanisns and pre-processing techniques can be directly applied to the user defined objects.• Since the user defined objects are geometric, computational geometric algorithms can be applied on them. There are some disadvantages of direct geometric representation of all user objects.1. Many spatial objecta (such as a river system) that should have simple one-stroke conceptual definitions result in complicated potentially incomplete definition in terms of geometric entities.2. Many conceptual objects can be defined not only by spatial objects but by some topological restrictions on such objects. A political map may need the inherent semantics of political system.3. Operations semantically associated with spatial objects do not have obvious geo- metric counterparts; for example, roads which are polylines can be traversed, but boundaries which are also representable by (closed) polylines are not traversable. As a matter of fact, traversal has more to do with users perception than geometry. Such restrictions are not apparent in geometric representation.There has been object-oriented approaches involving proper inheritance and opera- tion encapsulation. But that they are far from an adequate representation follows from these observations.• They maintain the basic premise that the user needs to explicitly define the geo- metric character of the objects.• The acknowledgment of the fact that only inheritance and encapsulation are not sufficient tools to model space. Relations other than IS.A occur in space and comes out from the mathematical structure of spatial domains. So, a semantically stronger model is necessary to abstract space.In this dissertation, we present a more conceptual data model developed in course of our work on an Archaeological Information System. This model provides a topological abstraction of geometric information and is borne out of a necessity to disassociate the semantic representation from the internal representation. Our query language allows processing of queries at this higher level of abstraction. The inner geometric structures are transparent and would be used only if the user's query cannot be answered at the semantic level. The complimentary use of the logical structure provides a more efficient utilization of the user's conceptual model of space. A second significant aspect of the model is that it builds on theme integration-which is akin to view integration in relational models with an important difference. A set of conceptually connected spatial objects are grouped into a structural aggregate called theme. Different themes may have different models and integration is affected through a shared name space and appropriate geometric anchors. The integration helps to answer intertheme queries in which the nodes of a query graph and the response belong to all different themes. The organization of the dissertation is as follows. In Section 2, we deal with the formal aspects of data model. Section 3 deals with the definition, specifications and operations on themes, our major focus of the dissertation. In Section 4 our query language and its processing are described. Section 5 illustrates strategies for processing inter-theme queries. In Section 6, we conclude with a discussion on our implementation and the scope for future work.

Comments

ProQuest Collection ID: http://gateway.proquest.com/openurl?url_ver=Z39.88-2004&rft_val_fmt=info:ofi/fmt:kev:mtx:dissertation&res_dat=xri:pqm&rft_dat=xri:pqdiss:28843301

Control Number

ISI-DISS-1988-189

Creative Commons License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

DOI

http://dspace.isical.ac.in:8080/jspui/handle/10263/6354

This document is currently not available here.

Share

COinS