The Preliminary Study of Chinese Gazetteer Information System
I-Mei Hung, Academia Sinica
yimay@pluto.ihp.sinica.edu.tw

Designing Gazetteers for Frequently Changing Places: Examples from Chinese History
Ruth Mostern, University of California, Merced
rmostern@ucmerced.edu

Extracting Geographic Features from the Internet to Automatically Build
Detailed Regional Gazetteers
Daniel W. Goldberg, John P. Wilson, Craig A. Knoblock, University of Southern California
dwgoldbe@usc.edu, jpwilson@usc.edu, knoblock@isi.edu

The Construct and Development of Gazetteer Database in Taiwan
Wen-Rong Su, Hsiung-Ming Liao, Yan-Tong Chen, I-Chun Fan
spidersu@gate.sinica.edu.tw

Using XML for Gazetteers and Other Place-based Information and
Cataloging
David Germano, University of Virginia
dfg9w@virginia.edu

Abstracts

The Preliminary Study of Chinese Gazetteer Information System
I-Mei Hung, Academia Sinica

Discussions of administrative place names in China involve spatial-temporal transitions throughout the past dynasties. These place names reflect specific political, economic, and social attributes of that time. As time passes by and the environment changes, an understanding of administrative place names is helpful in contextualizing historical knowledge. Therefore, to reflect the truth of historical records, historical studies demand an intersection of both temporal and spatial axis’s.

For a long time, major developments in Library Information Science have been in gazetteer, geographical heading and authority control. They mainly focus on temporal sequence and descriptions of property and attribute of place names, yet lack spatial concepts and spatial-temporal definition. On the other hand, major developments in geospatial information have been in GIS techniques, which focus on presenting the spatial concept of a specific period, but lack temporal sequence and analysis of place name characteristics and attributes. This paper, from the perspectives of these disciplines, intends to integrate the demands and approaches across disciplines, and to utilize metadata to construct a model of the administrative place names in China, which acts to organize and systemize the four-dimension variable structure (4D Structure) of administrative place names in China. At the same time, the application of GIS techniques can provide users with integration mechanism of gazetteer and thesaurus.

Designing Gazetteers for Frequently Changing Places: Examples from Chinese History
Ruth Mostern, University of California, Merced

Digital gazetteers are an invaluable tool for historical geography. At a minimum, index place names to feature types and locations. Fully-fleshed, standards-compliant gazetteers also relate alternative place names to one another, depict relationships among places, and describe spatial changes over time. Gazetteers can be incorporated into information systems that include map visualization, library catalogue searches, and links to additional information about places including other kinds of databases such as biography and bibliography.
Historical gazetteers can illustrate and structure the unstable and evolving ways that people name places and occupy space; however, their ability to do this has not yet been fully explored, and many problems remain. This paper uses examples from Imperial China to evaluate the capacity of gazetteers to depict the complexity of temporal change:
· Buddhist mountains, peaks, temple complexes and their constituent parts
· The intersection between the formal administrative system and the fanzhen system of military government during the Five Dynasties era
· Very frequent and provisional changes in administrative geography during the Song era.
This paper further asks whether developing historical gazetteers is a valuable use of time and resources. I argue that the labor of database development “pays off” by enriching knowledge about historical geography, creating a basis for information systems, and improving gazetteer standards.

Extracting Geographic Features from the Internet to Automatically Build
Detailed Regional Gazetteers
Daniel W. Goldberg, John P. Wilson, Craig A. Knoblock
University of Southern California

The utility of every imaginable application built using a gazetteer hinges on the simple fact that the resulting system will only be as useful, complete, or accurate as the underlying gazetteer itself. A major issue confronting gazetteers utilized in systems today is that they are not complete, and measures of their accuracy are largely unknown.

In this paper we describe an algorithm which addresses this problem by automatically generating highly complete and detailed regional gazetteers from internet sources. We utilize information extraction and integration techniques to automatically obtain geographic features and associate footprints and feature types from freely available online data. The sources used by our algorithm are widely available, and could be applied to create a gazetteer for any area. To evaluate this new automatically generated gazetteer, we extend previous work by defining several measures which can be used to assess the completeness and accuracy of gazetteers. Using these measures, we compare our gazetteer against existing gazetteers of two types; the Alexandria Digital Library (ADL) Gazetteer, considered to be the most complete gazetteer for the United States, and the Los Angeles County Bibliographical Database (LACBD), considered to be the most detailed gazetteer for Los Angeles County. Additionally, we discuss some defining characteristics of gazetteers created using our methods. Our results indicate that a gazetteer created by our methods will be at least as complete as any gazetteer currently available for certain types of feature classes, while falling short in others. We conclude by offering suggestions to address these shortcomings.

Geographic names are proper nouns that people use to designate a place or an area. Geographic names are the most common spatial reference and foundation of Indirect Geo-Referencing. They also represent background meaning of the environment. Most toponomy in Taiwan is limited to area studies which often emphasize the historical evolution of geographic names at village level and do not cover all geographic names in Taiwan, Ministry of the Interior and different map-making departments have established plentiful gazetteer database. However, a lack of standardization not only hinders the development in toponomy but also causes inconvenience of the public. A standardized national gazetteer database will contribute to a more systematic map making and more efficient land administration. This research project surveyed geographic names throughout Taiwan to collect historical and etymological information. We used computer database technique to construct the Taiwan Gazetteer System, which allows users to retrieve information about geographic names online. We have collected approximately 13,100 records of geographic names and the system has been open for online service. This project is based on the ADL Gazetteer Content Standard, which is developed by the Alexandria Digital Library Project, to define the gazetteer standard of Taiwan. We also followed the description of ISO Standard and modified some classification, codingetc to fit in with the conditions of Taiwan. Key words: Gazetteer, Standard, Geographic name 1 Assistant of GIS Program, RCHSS, Academia Sinica 2 Chief of GIS Program, C.C., Academia Sinica 3 Associate Research Fellow, I.C.H., Academia Sinica

Using XML for Gazetteers and Other Place-based Information and
Cataloging
David Germano, University of Virginia

I will present the details of a Gazetteer DTD that has been created by the University of Virginia Library's Digital Library Research and Development Department in conjunction with the Tibetan and Himalayan Digital Library. The presentation will focus on technical problems, and their resolutions, in our struggle to make an XML DTD adequate to the challenges of creating a Gazetteer. The presentation will begin by discussing issues pertaining to the choice of XML rather than MYSQL, Postgres, or other types of traditional database technologies. In this regards, I will address XML-specific problems we continue to struggle with, such as indexing-searching, and workflow management including communal editing. XML is well suited for working in fluid fashion with hierarchical data, but can be cumbersome in terms of its
creation and editing tools. In addition XML carries its structure with the data, and is excellent in dealing with repeatability and recursion. These and other issues will be examined in relationship to Gazetteers and their needs.

The rest of the presentation will then look at specific problems we have struggled with in the DTD: how to keep track of alternative names and different linguistic forms of names, how to track and express relationships between places and especially hierarchical relationships, and other such issues. I will also discuss our strategies for using this Gazetteer as a service to georeference media collections using the feature IDs- images, audio-video, texts - and the challenges we face in that regards. Finally, I will address our strategies for using a separate DTD - GDMS – for representing richer data about places.