From: rick@cs.arizona.edu (Rick Schlichting) Subject: APPS: Kahaner Report: Geographic Info Sys; Large Spatial DBs, Singapore 6/93 Date: 23 Jul 93 12:15:46 GMT Organization: University of Arizona CS Department, Tucson AZ Crossposted from comp.research.japan [Dr. David Kahaner is a numerical analyst on sabbatical to the Office of Naval Research-Asia (ONR Asia) in Tokyo from NIST. The following is the professional opinion of David Kahaner and in no way has the blessing of the US Government or any agency of it. All information is dated and of limited life time. This disclaimer should be noted on ANY attribution.] [Copies of previous reports written by Kahaner can be obtained using anonymous FTP from host cs.arizona.edu, directory japan/kahaner.reports.] From: Dr. David K. Kahaner US Office of Naval Research Asia (From outside US): 23-17, 7-chome, Roppongi, Minato-ku, Tokyo 106 Japan (From within US): Unit 45002, APO AP 96337-0007 Tel: +81 3 3401-8924, Fax: +81 3 3403-9670 Email: kahaner@cs.titech.ac.jp Re: Geographic Information Systems; Large Spatial DBs, Singapore 6/93 21 July 1993 This file is named "gis.93" ABSTRACT. The 3rd International Symposium on Large Spatial Databases (SSD'93), and the Far East Workshop on Geographic Information Systems (FEGIS'93), Singapore June 1993 are reviewed. Some additional remarks about computing activities in Singapore are included. The week 21-25 June 1993 was devoted to Geographic Information Systems (GIS) and associated database technology in Singapore. A two-day workshop on GIS, the Far East Workshop on Geographic Information Systems (FEGIS'93) was held at the plush Oriental Hotel with the assistance of local corporate sponsorship. Following, the three day international conference, Large Spatial Database (SSD'93) was held on the campus of the National University of Singapore. GIS are essentially computerized decision support systems which involve the integration of spatially referenced data. Applications are wide and multi-disciplinary, such as cartography, environmental protection, surveying, hydrology, forestry, urban & regional planning, transportation planning, navigation, etc. Geographic data includes maps but also many other objects and relations about objects in space. One general definition of GIS (O'Callaghan) gives it as a system for capturing, manipulating, analyzing and displaying data which are geographically-referenced to the Earth. Here, geographic refers to human scale, i.e., mountains rather than molecules, in which the user envisions moving around within the data rather than moving the data objects. Early GIS concentrated on capabilities for data capture, manipulation and mapping in the general sequence, capture-->manipulation-->analysis-->display. There has been, and continues to be a significant interest in GIS within the military community, and it is clear that the parallel processing implementations for rapid military intelligence are being pursued in several countries. However, GIS has also been heavily associated with database (DB) technology because of the large volumes of data that are traditionally associated with geographic information. Early GIS incorporated separate facilities for managing geographic data and attribute (either numeric or factual) data. The early GIS systems were built on existing relational database packages. There is still research in the area of building sensible GIS query systems that then are converted to SQL queries on ordinary DB systems, or developing such queries using conventional DB languages. More recently, there has been a trend to merge all types of data with the use of more flexible DB systems, in particular with Object-Oriented databases (often using the C++ language) which seem extremely well suited to geographic issues. These have the potential to include multiple formats (video, image, graphics, voice, and text). Query capabilities are also improving, with much less structured queries and the ability to support browsing, navigating, etc. As DB capabilities have improved so have GIS. Also, as capture (digitizing) and storage of geographic data has become more automated, there has been more time and motivation to focus on the high-end parts of the process, as well as a trend toward making them less sequential and more interactive. (One SSD'93 participant commented that currently 90% of GIS costs are related to data generation and gathering, and that much more work needs to be done on user interfaces which remain weak.) Better user interfaces including multi-media and interaction, as well as more sophisticated and faster DB storage and query techniques have immediately been put to use by the GIS community (in fact many GIS researchers could be classified as DB researchers and conversely). Although end user requirements are not well understood, the current trend is to move GIS from a specialist tool used by a technical group to an integrated decision making tool that will be used for both spatial and temporal data management of complex phenomena at various levels within organizations including operations, logistics, and executive. In other words, to become part of information systems services, active document (paperless) systems, enterprise systems, etc. In particular, as industry becomes more distributed and global, the use and need for geographical information is bound to increase. (Several speakers mentioned industrial applications related to tracking of equipment, packages, personnel, etc., by using GIS in conjunction with global positioning systems.) GIS are not as widely used in Asia as in the West. However, over the past few years many large GIS projects have been initiated as the economies of these countries pick up steam. Every country that I have visited in Asia has a very high level of interest in GIS. Usually there are major projects supported at the top levels of government. The potential applications of GIS are easy to describe and demonstrate and invariably enthuse both politicians and bureaucrats who can imagine the benefits. Because many Asian countries are just now beginning to grapple with problems of agriculture, planning, environmental assessment, etc., GIS are seen as a potentially invaluable tool. Intergraph, a company which provides software and related GIS services, illustrated examples of their products in use in Singapore, Hong Kong, Australia, Philippines, New Zealand, Thailand, etc. Some fundamental research in GIS and related DB topics is occurring in Asia, but many of the papers presented at the FEGIS'93 workshop were detailed applications of GIS to specific city, country, or regional needs. Such examples were provided by speakers from Indonesia, Pakistan, Taiwan (oceanic info system), Australia, Bangladesh, Malaysia, Vietnam, Singapore, etc. Thus the workshop provided many "case studies". One very interesting short paper showed the variety of possible applications for satellite data in China and the need for integration and standardization. The workshop thus provided opportunities for people who wanted to learn about GIS, for practitioners to describe specific implementation/development projects that they were working on, and for vendors to display products at the associated exhibit. A few papers were clearly marketing opportunities. (I have no objection to such talks as they often convey useful information, but feel that the speaker's affiliation and position should be clearly indicated in the program to allow participants to make sensible judgements--especially in the case of parallel sessions.) While many papers would not be classified as pure "research,", some were very interesting nonetheless. In the following I have summarized several of the papers that were of special interest to me. A complete list is attached to this report. In addition, the proceedings have been published as follows. GIS: Technology and Applications Proceedings of the Far East Workshop on Geographic Information Systems Eds: Hongjun Lu & Beng Chin Ooi World Scientific Publ Co Pte. Ltd PO Box 128 Farrer Road, Singapore 9128 (US Address: Suite 1B, 1060 Main St., River Edge NJ 07661) ISBN 981-02-1445-6 (1993) In terms of general vision (within the Asian community) I should point to remarks by Australia's CSIRO Division of Information Technology under John O'Callaghan, whose keynote speech clearly showed that GIS will spread toward organization-wide services. He gives a simple example of a water utility which "could use the DB of its reticulation network to support not only planning and design of the network but also activities concerned with network maintenance, emergency services and marketing. Information on these activities could be made available as map-based documents, able to be interrogated as appropriate throughout the utility." CSIRO appears to be one of the largest Asian centers of research in GIS and related technology. Callaghan's IT Division has more than 100 staff with a budget of about A$9M at three sites (Sydney, Canberra, and Melbourne), with 20 professionals working at the Canberra lab on Spatial Decision Support Systems (distributed databases, systems integration, real time spatial information systems). [Email: JOHN.OCALLAGHAN@CSIS.DIT.CSIRO.AU]. Two very interesting research papers were given by Singaporians. Prof Beng Chin Ooi from the National University of Singapore [Email: ooibc@iscs.nus.sg] described fundamental work in knowledge discovery (done jointly with colleagues at Simon Fraser University. Dr. Jian Kang Wu [Email: JIANKANG@ISS.NUS.SG] at the Institute of Systems Science described a large knowledge-based GIS that he designed and implemented while at the the University of Science and Technology in PR China based on remote sensing data. This was the first GIS software system in China and we were told that this is now in the stage of product development. There were a number of papers concerned with implementing O-O ideas into GIS typically with implementation in C++ (ex, Wang & Lee), and data structures for dealing with the special characteristics of geometric data (ex, Kim & Bae). Sometimes data structures and user interface are considered together (ex Nakamura, Abe, & Ikeda). Position uncertainty is a common problem (digitized maps or platt plans often have errors)--Shibasaki discusses the relationship between data structures (point, line, polygon) that allow for such errors using a probabilistic approach in one paper and fuzzy inferencing in another. Several speakers commented on the use of real-time positional satellite information for navigation and to reduce surveying errors, and Frank even suggests that future surveyors will wear VR type glasses; when an object in the glasses (computer data) appears to be be out of place relative to its physical entity (a pipe is displaced a few meters), the surveyor will use a data glove to move the computer entity to correspond to its proper position, and the GIS will appropriately update the internal data representation. Frank points out that it takes 7-8 years for research lab results to get into the marketplace, and thus he views the turn of the century as the most likely time for today's lab work to be commercially viable. Or to say this in another way, "very little will reach the market in 10 years, which is not yet studied in a research lab." With O'Callaghan, he feels that GIS will become more integrated into business systems and that communications will be the key technology to make that happen. Integration is already occurring in some places. Madhav (Singapore) describes integrating GIS with workflow management. Also see my remarks below about Singapore's data hub project. The Large Scale Database conference (SSD'93) which occurred immediately after FEGIS'93 was essentially an international conference that just happened to be held in Singapore. By that I mean that there were only two or three papers from Asian institutions, Singapore, Taiwan, and Australia. Generally, Singaporians were listening and learning. A widely available Proceedings has been published, Advances in Spatial Databases Third International Symposium, SSD'93, Singapore, June 1993 Eds: David Abel & Beng Chin Ooi Lecture Notes in Computer Science # 692 Springer Verlag, Berlin ISBN 0-387-56869-7, 1993. SSD'93 had about 170 participants from all major countries as well as from several smaller ones (ex: Saudi Arabia and Lybia). A list of titles and authors is attached to this report. I will concentrate on only two points that I felt were particularly interesting, allowing readers to refer to the Proceedings for details on the other international contributors. At the opening of SSD'93 there was a speech by Mr Tan Chin Nam, chairman of Singapore's National Computer Board, and managing director of the Economic Development Board. I felt that his remarks were particularly insightful and have included large sections below, with the caveat that there may have been some alterations in his speech that I missed. Please note particularly the four items considered to be essential for GIS in Singapore -- support for distributed data-bases, standards for data exchange, ease of use, and training. "Whilst spatial information technology has been applied in many areas such as CAD/CAM, multi-media, imaging, visualization, one of the most promising applications is Geographic Information System. I would like to make some comments on this particular application. Singapore, as you know, is a small island of about 640 square kilometer with about 3 million people. Land is a precious resource. Singapore's physical and intrastructural development is driven by the needs of a growing economy and, more and more, by the changing aspirations and expectations of an increasingly affluent society. Hence it is of critical importance that we do the best job in managing the usage of our land resources, implementation of telecommunication and utility lines, construction of roads and buildings, and urban planning. That is why we have begun to exploit geographic information systems. In Singapore, GIS is widely used in the public sector. The Integrated Land Use System (ILUS), Singapore's first large-scale GIS project, is being developed to support the city planning, development and building control, and road planning functions of the Ministry -of National Development. The Public Utilities Board is using GIS to manage the utility lines, Singapore Telecom to manage the telecommunication lines and the Ministry of the Environment to manage the drainage and sewerage system. The Ministry of Law uses the technology to manage the legal boundaries of land parcels and to produce the Singapore street directory. A key success-factor of these GIS projects is the availability of relevant data. Often, geographic data collection is an enormous task requiring substantial effort because the data reside in many different government agencies. In Singapore, a Land Data Hub was set up in 1990, to administer land data and coordinate land data sharing in the public sector. Using GIS, the Land Data Hub integrates vast volumes of data scattered among 12 public organizations. With the hub, a wealth of spatial and textual data can be exchanged and shared. Currently 23 Government departments and statutory boards subscribe to the services [see also remarks about this hub below, DKK]. GIS applications are not restricted to land management, urban planning, facilities management, automated mapping and resource management. An application which could make good use of GIS is site selection by retail outlets. Studying market profiles in a certain locality could help a businessman decide on the actual site of a retail outlet, products to be carried and the kind of targeted promotion programs to launch. More applications could be made in the transportation sector. With GIS, we could have a smart road infrastructure that would permit on-board vehicle navigation. Such a system could help us, as drivers or passengers, to select the best route to the office or other places of interest, taking into account on-the-spot traffic conditions. Businesses that move goods on land can reduce their cost by optimizing their distribution routes. In an emergency, traffic could be diverted automatically and routes cleared for emergency vehicles. The minutes saved could mean a difference between life and death. The potential of GIS is vast. Even President Bill Clinton used GIS to help him determine how his campaign resources could be best deployed in the 1992 presidential election. His staff used GIS extensively to show voting trends and voter profiles throughout the country, identifying areas where elections had been close in the past. The information helped the campaign tremendously in targeting media advertising and determining where to send the candidates. It even helped with the drawing of the Clinton-Gore bus route through the Midwest after the Democratic convention. This resource targeting strategy worked and also helped save money for the 1992 Democratic presidential campaign. The opportunities for using GIS are plentiful and the benefits to be reaped from it appear to be tremendous. However, for the technology to be exploited on a large scale, four issues need to be addressed. Firstly, there is a need to support distributed spatial databases. Often, data that support the spatial analysis of business functions are stored in different databases at different locations. Open standards for heterogeneous spatial database environments need to be established so that the databases can 'talk' to one another. The second issue concerns establishing a national or even an international graphic interchange standard for effective and efficient data exchange. As spatial information technology is multi-disciplinary, the data tend to be different in formats and models. A common data exchange format and model for the entire spectrum of users does not fully exist yet. In Singapore, graphic interchange standards is one of the priority areas which our national standards committee is actively looking into. The third and an important issue is the ease of use of the system. No matter how powerful the computer hardware and software are, it means little if the average office worker or in the case of a public transport system, the average commuter, cannot use it. Also, it needs to be made affordable and as commonplace as personal computer word-processing and spreadsheet packages. Finally, we need a workforce that is trained to use spatial technology. I understand that the [Singapore] National Institute of Education has acquired GIS packages to teach geography. Also, I have been told that most current computer science curricula already include courses on computer graphics, database applications and multi-dimensional data structures at the undergraduate level. However, it would be desirable to have more advanced courses, perhaps at the postgraduate level, to train specialists in spatial information technology. I would like to take this opportunity to urge the government organizations, spatial technology experts, and the industry end-users to work together to help accelerate the development and exploitation of this powerful tool." An invited talk was given by Ms Yuk-Wah Tang-Kwong, who is the manager of the 17 person Land Systems Support Unit, the operational arm of the Singapore Land Data Hub. The outlines of this were described above. Here I will add that plans are for a truly national integrated DB. Work on the hub began in the early 1980's with existing DBs for land, and people and has been significantly expanded (although still not complete). About 92% of the cadastral land lot boundaries are included and the remainder are scheduled to be added by the end of 1993. Gas, water, electricity, drainage and sewer networks have been imported into the hub. The hub will eventually contain proposed and existing graphic outlines and related textual data for buildings constructed by both government and private contractors. Three of four agencies responsible for road construction have supplied road outlines to the hub, with road center line information based on 1:5000 topography maps. These maps, which also show roads and tracks, outlines of buildings and other structures, hydrographic features, footpaths, etc., are also mostly entered into the hub. Other categories include the Singapore Street Directory, lot base information, government building data, etc. The obvious benefits, as mentioned, are integration, consistency, and quality improvements in existing services, with opportunities for new applications. Institutional, rather than technical issues have proved to be the major challenges, although data sharing between organizations has not been a problem. Large amounts of effort are expended on fixing discrepancies, vertical and horizontal integration (does detailed data from one source agree with higher level data from other sources, etc), data harmonization, and related quality aspects. There is a central copy of common sharable data (although providers have their own copies), a corporate data manager for spatial and non-spatial data, semi-automated updating and map replacement, etc. Major challenges are related to the comments by Mr Tan Chin Nam, ease of use, distributed model, low cost, as well as the additional item of data security. The next step is to merge the hub with telecommunications data exchange services to further integrate Singapore's data bases (100% of Singapore's telephone network information is now being digitized, and will probably be the first to be 100% fiber to the home). (One Western participant asked "how did you get people to agree to share data -- we have many difficulties with this." The speaker admitted that she did not know why it was easy to get cooperation--my own feeling is that she should have responded rhetorically, "why is it so difficult for you, when the benefits for all are so clear?") One unusual paper (Williams & Woods) on expectation in spatial information was related to the problem of building a tactical military intelligence processing system. An interesting problem, yet many attendees were clearly uncomfortable with the topic. M. Stonebraker (CS Berkeley) described the objectives of the SEQUOIA 2000 project, which has the goal to build a better computing environment for global change researchers, a topic that he felt was the key science issue for our generation. The paper presented at SSC'93 (included in the Proceedings) is an excellent readable summary of the plans. The project has substantial funding from a variety of companies (including DEC, HP, etc), US science agencies (ARPA, NSF, ARO, NASA, USGS), as well as funding from California's Dept of Water Resources. A key aspect of this project was the performance required -- high performance I/O on terabyte data sets, high speed networking, powerful visualization tools, etc. I mention this because I was surprised that during the five days of papers at both SSD and FEGIS, large scale performance issues did not seem to play a major role. My sense is that, especially in Asia, the emphasis is on GIS systems that run on workstations (mostly singletons, but a very few distributed systems) and PCs. The typical organizations that are using GIS in this part of the world do not yet have access to really powerful computational platforms. Nevertheless with the availability of high performance workstations and servers, GIS can now be implemented using a client-server approach. In this, the server handles the data retrieval and storage, while the clients run individual application programs on their workstations. SSD'93 ended with a panel addressing the question "Are We on Track?". Stonebraker presented a list of issues that needed to be worked on as well as those that (he felt) were settled. Work needed on: Visualization tools Abstracting Lineage Uncertainty in data Object oriented GIS and DB design tools Data mining Integration of visualization, DBMS, models (for steering) Classification functions DBMS extension mechanisms Multi-type operators (such as joins) over heterogeneous types and DBs Work not needed on: Spatial access methods Geographic SQL The topic of spatial access provided the only area where there were major disagreements among participants. Spatial access refers to data structuring techniques for describing complicated geographic data. The key issue here is rapid access, and a variety of tree-traversal techniques have been developed (R trees, quadtrees, pyramid, BSP trees, K-d trees, etc). Several variants were presented at either the workshop or SSD'93. One of the world's experts in this topic is H.Samet (CS U-Maryland), and he strongly disagreed with Stonebraker's view that enough had been done. Samet remarked that many GIS users are in government organizations where costs and efficiency are not a major issue. But within industry the situation is very different. Companies using efficient access methods are reluctant to talk about this for fear of giving away an economic advantage. He commented specifically on a parcel delivery service in the US which was vitally interested in efficiency. Other points made during the panel were similar to those made during individual papers, the need to interoperability, better visualization, scale transformations, growth of O-O technology, etc. There was also the re-emphasis on the need to educate users and several comments that for GIS uses, political issues dominated technical ones. In response to questions about how additional research funding could be generated, Stonebraker and other "successful" researchers remarked that it was important that GIS researchers had "customers" whose needs they kept in mind, as opposed to traditional academic research projects. As an indication of the need to link to other scientists, SSD'93 participants were reminded that there were no interdisciplinary papers presented. ---------- Singapore: Additional remarks. As I have written in several earlier reports, Singapore is hard at work trying to make itself the technology capital of the Pacific rim--the Intelligent Island concept. The computer part of this is called IT2000, which is being implemented by the National Computer Board. I will have more details about the general plans in the future, but here I should mention some progress that was evident even during a brief visit. The second phase of the country's science park has just opened, and should be completed by 2001. (This was modelled on St John's Innovation Center in Cambridge University, UK.) The first tenants are to be the Institute of Microelectronics and the Information Technology Institute. The existing section has a variety of R&D organizations, including many government agencies. There are some unusual ones too; Reuters software occupies one of the largest and most elaborate building within science park. Singapore is making a major effort to be the conference center of the region. English language conferences are in progress virtually nonstop all year; I cannot keep track of them all, and could only attend a very small fraction. This is a friendly and effective way to learn about advanced technology as well as a mechanism for sharing technical ideas. To the extent that Singaporian participation is mostly passive, information flow is basically inward, but as the level of local expertise increases, such conferences are useful for all attendees. Conference fees and hotel charges are relatively high, so this is probably a significant commercial activity too. -------- NUS CS & MATH DEPTS In addition to hosting SSD'93, many CS members were participating in the conference which was held in the same building as most Department offices. I had a brief introduction to the Department's activities by Prof H.W.Leong Dept of Information Systems & Computer Science National University of Singapore Lower Kent Ridge Road Singapore 0511 Tel: +65 772-2734; Fax: +65 779-4580 Prof Leong (PhD Illinois) works in the general area of fundamental algorithms with specific application to VLSI design automation. He is coordinating the Department's Algorithms and Design Automation Lab which has six faculty members and seven junior faculty (some faculty also belong to other research labs). Very recent papers (jointly with a PhD student) include the following. Probabilistic Analysis of Memory Reconfiguration in the Presence of Coupling Faults, C.P.Low & H.W.Leong, Proc IEEE Intl WS on Defects & Fault Tolerance in VLSI Systems, 4-6 Nov 1992, Dallas TX. Efficient Algorithms for Regonfiguring Memory Arrays in the Presence of Coupling Faults, C.P.Low & H.W.Leong, Proc IEEE Asia-Pacific Conf on Circuit and Systems (APCCAS'92), 8-12 Dec 1992, Sydney Australia. On the Reconfiguration of Degradable VLSI/WSI Arrays, C.P.Low & H.W.Leong, Submitted to DFT'93, Venice Italy. Some of his algorithms have been implemented on an RS/6000 department workstation. There seemed to be an adequate number of workstations and related facilities for general purpose research. I also met briefly with Prof Chong Chi Tat Vice Dean of Science & Professor of Mathematics National University of Singapore Lower Kent Ridge Road Singapore 0511 Tel: +65 772-2833, -2752; Fax: +65 777-4279 Email: SCICCT@NUSCC.NUS.SG to discuss the status of NUS's computational science program. I wrote about this just as it was beginning about two years ago. Prof Chong remarked that the first group of students would graduate soon, so it was still too early to determine the actual success of the program. However, enrollment was increasing and the program seems to be well accepted. Prof Chong, with other scientists in Singapore are now working on a plan to upgrade the high performance computing capabilities available (see below) to the country's S&T community. (There is a significant amount of engineering and other activity in Singapore that could make good use of high performance computers. Currently there is an SX-1 that clearly needs to be either replaced or updated.) It is my opinion, which is shared by many in Singapore, that significant additional capacity is seriously needed. Without it, scientists will naturally restrict the kinds of problems they tackle, which alters their worldview. It is a feedback loop that is counterproductive. INSTITUTE OF SYSTEM SCIENCE While SSD'93 was at the National University I also visited the Institute of System Science (ISS) which is across the campus. I have reported on this before but my return visit reaffirmed that this is an impressive place which is getting significantly stronger. ISS is working in the information technology arena, with particular emphasis on four topics. Human interface technology Machine intelligence Networking and information architectures Parallel architectures and algorithms There are subtle, but real differences between research at ISS and that conducted at NUS. This can be summed up by making the analogy between a Western university CS department and an associated university laboratory. ISS's research is project oriented and managed by project managers with activities grouped by teams, and explicitly intended to fulfill the vision proposed by the country's National Computer Board toward making Singapore a more high tech community. ISS has a combination of talented staff, excellent facilities (more than 100 Unix workstations and associated equipment housed in a new US$15M building) in an on-campus location, project oriented research, good links with industry (Apple and IBM are conspicuous), in a fully English language environment which is relatively open. In this report it is not appropriate to describe research here in detail except to mention several new staff. Dr. Raghu Raghavan a Wisconsin PhD in Physics who has worked in the US in automata, pattern recognition and real-time parallel computing is leading work on multi-computing over high speed networks. I have invited Dr Raghavan to participate in the Supercomputing Around the World Minisymposium to be held as part of Supercomputing '93 in Portland Oregon (Nov 1993), where he will describe Singapore's plans and projects for high performance computing. [Email: RAGHU@ISS.NUS.SG]. Dr Timothy Poston, a Warwick Phd in Mathematics who has written texts in Differential Geometry and Catastrophe Theory, is working in medical imaging issues related to computer vision, neural networks, the geometric structure of the human brain and also working on human computer interfaces. I met Dr Poston first in Korea; an eclectic, he has published in physics, biology, mathematics, and archeology [Email: TIM@ISS.NUS.SG]. Dr. Wu Jian Kang, a U-Tokyo PhD in computer science, is also full professor of the University of Science and Technology of PR China. I mentioned his work on GIS earlier (see above). Dr Wu has recently been working on a system for facial image retrieval with identification and inferencing capabilities using both neural nets and fuzzy sets, and will present a paper on the topic at the ACM Multimedia Conference in Anaheim CA in Aug 1993. ISS will host the first Asian Fuzzy Systems Symposium 23-26 Nov 1993. It was also one of the hosts for SSD'93. ---------------------------------------------------------------------- Program: Far East Workshop on Geographic Information Systems, 21-22 Jun 1993, Singapore KEYNOTE SPEECH: The Impact of Applications and Information Technologies on Geographic Information Systems, J.F. O'Callaghan SESSION 1: GIS Technologies: Data Handling and Models Chair. Chandra Bhople, SysScan Asia Pte Ltd, Singapore Models for the Management of Parcel Data (invited talk), J. H. English, USA Extensions to Object-Oriented Data Models for GIS Applications, C. Lee and S.-C. Wang, Taiwan A Framework for Handling Geometric Data with Positional Uncertainty in a GIS Environment, R. Shibasaki, Japan Rr-Tree: The Design of an Efficient Access Method for Spatial Objects, J. Kim and H. Bae, Korea SESSION 2: GIS Applications: Urban Planning Chair.- William Shepherd, ESRI, Singapore Multi-Purpose GIS/LIS as Applied in Housing and Development Board of Singapore, T.K. Tang and Y. K. Tiong, Singapore A System for the Evaluation of Urban Environment, W. Yan, H. Nakamura and R. Shibasaki, Japan The Use of GIS to Monitor Urban Development, Case Study.- BOTABEK Region, Indonesia , Ir. Tubagus Haedar Ali, Indonesia Urban Database of Karachi to Strengthen the Planning Process A. Hasan, A. Baig, A. Raza, and N. limer, Pakistan TUTORIAL 1: Spatial Query Language for GIS, Max J. Egenhofer, USA SESSION 3: GIS technologies: System Design and Performance Chair. Clement Leung, NUS, Singapore A Comprehensive Approach to GIS CostIBenefit Analysis (invited talk), A. McLachlan, Hong Kong Interactive Graphics and Spatial Data Management for GIS Using the Hierarchical Data Structure, Y. Nakamura, S. Abe and K. Ikeda, Japan Geographical Information Systems Development Using a General- Purpose Database Programming Language, T.H. Merrett and B. Khalffe, Canada Database Design and Implementation of Oceanic Information System, S.-T. Lee, Taiwan SESSION 4: GIS Applications: Land Information Systems Chair. Tang-Kwong Yuk Wah, NCB, Singapore The Australian Geographic Database Program. Geodata (invited talk), A. Clarke, Australia Electronic Round Table: A Contribution to Cooperative Land Management, R. Lake, Canada Land Information Systerm For AsiA K. Kubik, Australia A Large Scale Land Information System (LIS) for the Structural Analysis of Agriculture in Flanders (Belgium), J. Desmet, Belgium KEYNOTE SPEECH: GIS in the Year 2000, A. U. Frank (Austria) SESSION 5: GIS Technologies: Query Languages and User Interface Chair. Hans-Peter Krieger, University of Munich, Germany Designing a Spatial Query Language for Spatial data Query and Analysis, Z. Huang, Sweden An Experimental Investigation of Color Selection for Digital Street Maps: Implications for User Interface Design, A. H.C. Lim and E.-H. Chia, Singapore SESSION 6: GIS Applications: Risk Assessment Chair. Dave Abel, CSIRO, Australia Delineation of Catchment Area and Proximity Analysis of Cyclone Shelters in Coastal Region of Bangladesh Through GIS, F. Ren, M.R. Rahman, S. Ochi and S. Mural, Japan GIS and Acid Rain in Malaysia, K. L. Cheek and T. Buyong, Malaysia SESSION 7: GIS Technologies: Knowledge Engineering and GIS Chair. Son Dao, Hughes Research Laboratories, USA Keeping Pace with Technological Advancements - the Evolution of GIS Technologies (invited talk), D. Byers, USA. KGIS: A Knowledge Based Geographic Information System, J.K. Wu, China Discovery of General Knowledge in Large Spatial Databases, W. Lu, J. Han and B.C. Ooi, Canada SESSION 8: GIS Applications: Resource and Environment Management Chair. Taher Buyong, UTM, Malaysia Geographic Information Systems for Forest Resources, Wildlife, Environment Management in Vietnam, H. P. Lai, Vietnam Area Survey Photographic Satellite Applications in Different Fields of China, P. Ye and L. Li, China GIS Functionalities for the Analysis of Services Usage with Emphasis on Spatial and Socio-Economic Differentiation, P. van Teeffelen and T. de Jong, The Netherlands TUTORIAL 2: Towards the Next Generation of GIS, Michael Freeston, Germany SESSION 9: GIS Technologies: System Development and Education Chair. Chia Eng Hock, NUS, Singapore Which Concepts Should Form the Basis for a Geographic Information Management Systems (Invited Talk), K. Kjenstad, Norway A New Design of a Geogrlphic Data Supply Management System Applying Systematic Geographic Query Concepts, Tzu-How Chu, Taiwan Towards the Development of Low Level User-Oriented Geographical Information Systems, A. Evans, B. Garner and Q. Zhou,Australia Developing Vocational Training Institute for the GIS Industry, B. Leis and D. Mills, Australia SESSION 10: GIS Technologies: System Integration Chair. Jiawei Han, Simon Fraser University, Canada Information Environment in Utility Corporations, H. Nimtz, Germany Integration of Heterogeneous GIS for Environmental Planning and Management, C. Claramunt, M.H. de Sede, R. Prelaz-Droux and L. Vidale, Switzerland Implementation of GIS for Managing Very Large Databases in Development Projects, K.E. Seetharam, N. Hirose, A. Ishibashi, and S.Lai, Japan Integrating GIS Applications and Workflow Management Systems, R. R. Madhav, Singapore Program: Large Spatial Databases (SSD'93), 23-25 June 1993 Singapore Opening SSD'93 Chairman, C. K. Yuen (Nat'l. U. of Singapore) Keynote Paper Spatial data management in database systems: Research directions. Won Kim (keynote speaker), Jorge Garza, Ali Keskin (USA) Session 1: Data Modeling Session Chair: Per Svensson Realms: A foundation for spatial data types in database systems. Ralf H. Gueting, Markus Schneider (Germany) A canonical model for a class of areal spatial objects. Michael F. Worboys, Petros Bofakos (UK) Strong integration of spatial domains and operators in a relational database system. Thierry Larue, Dominique Pastre, Yann Viemont (France) Session 2: Spatial Indexing Chair: Hans-Peter Kriegel The transformation technique for spatial objects revisited. Bernd-Uwe Pagel, Hans-Werner Six, Henrich Toben (Germany) A paging scheme for pointer-based quadtrees. Clifford A. Shaffer, Patrick R. Brown (USA) Session 3: Indexing Mechanisms Chair: Hans Schek A hierarchical spatial index for cell complexes. Leila De Floriani, Monica Pellegrinelli, Elisabetta Bruzzone (Italy) On optimal multiversion access structures. Bruno Becker, Stephan Gschwind, Thomas Ohler, Bernhard Seeger, Peter Widmayer (Germany) Concurrent accesses to R-trees. Vincent Ng, Tiko Kameda (Canada) Session 4: Handling of Raster and Vector Data Chair: Max Egenhofer A spatial data model and a parallel topological sweep algorithm for map overlay. Ulrich Finke, Klaus Hinrichs (Germany) An optimal quadtree translation algorithm. Chuan-Heng Ang (Singapore) Database support for multidimensional raster data. Peter Baumann (Germany) Tutorial: Hierarchical Data Structures for Spatial Databases. Hanan Samet (USA) Keynote Paper From extensible databases to interoperability between multiple databases and GIS applications. Hans. Schek (keynote speaker), Andreas Wolf (Switzerland) Session 5: Spatial Database Systems Chair: Hanan Samet Interoperability of spatial and attribute data manager: a case study. Curtis P. Kolovson, Marie-Anne Neimat, Spyros Potamianos (USA) GEO2: Why objects in a geographical DBMS? Benoit David, Laurent Raynal, Guylaine Schorter, Veronique Mansart (France) Session 6: Topology Chair: Graham Williams A small set of formal topological relationships suitable for end-user interaction. Eliseo Clementini, Paolino Di Felice (Italy), Peter van Oosterom (Netherland) Qualitative and topological relationships in spatial databases. Z. Cui, A. G. Cohn, D. A. Randell (UK) Topological relations between regions in R2 and Z2. Max J. Egenhofer, Jayant Sharma (USA) Session 7: Storage Management Chair: Hongjun Lu Query-adaptive data space partitioning using variable-size storage clusters. Gisbert Droege, Hans Schek (Switzerland) A storage and access architecture for efficient query processing in spatial database systems. Thomas Brinkhoff, Holger Horn, Hans-Peter Kriegel, Ralf Schneider (Germany) Query processing of spatial objects: complexity versus redundancy. Michael Schiwietz, Hans-Peter Kriegel (Germany) Panel 1: Environmental data management - our great challenge. Chair: Oliver Guenther (Germany), Panelists: to be announced. Tutorial: Towards Intelligent Spatial Database Systems. Jiawei Han Keynote Paper The SEQUIOA 2000 project Michael Stonebraker (keynote speaker), James Frew, Jeff Dozier (USA) Session 8: Query Retrieval Chair: Ralf Gueting Neighbourhood query and analysis with GeoSAL, a spatial database language. Zhexue Huang, Per Svensson (Sweden) Application of a reciprocal confluence tree unit to similar-picture retrieval. D. J. Buehrer, C. C. Chang (Taiwan) Session 9: 3-Dimensional Data Handling Chair: Oliver Guenther Indexing on spherical surfaces using semi-quadcodes. Ekow Otoo, Hongwen Zhu (Canada) Volumes from overlaying 3D triangulations in parallel. W R Franklin (USA), Mohan Kankahali (Singapore) A declarative, object-oriented interface to a solid modeler. Michael L. Haytens (USA), Cristiano Sacchi (Italy) Invited Paper The GIS Projects in Singapore. Tang-Kwong Yuk Hwa (LSSU, Singapore) Session 10: Knowledge Engineering in SDS Chair: Mike Freeston Deduction and deductive databases for geographic data handling. A. I. Abdelmoty, M. H. Williams, N. W. Paton (UK) Representing expectations in spatial information systems: A case study. Graham J. Williams, Steven G. Woods (Australia) Panel Session 2: Are we on track? Panel Chair: David Abel (Australia) Panelists: Mike Freeston, John Herring, John O'Callaghan, Hans Schek, Mike Stonebraker ----------------------------END OF REPORT--------------------------