Integrating CHGIS with other Gazetteer Standards and Classification Systems Merrick Lex Berman, Harvard Yenching Institute PNC – ECAI, Guadalajara Dec 2001
CHGIS Organization sponsor research partners Harvard Yenching Institute Fudan University, Center for Historical Geographical Studies ACASIAN, Griffith University Academia Sinica – Taipei, Inst for Information Sciences
Tan Qixiang’s “Historical Atlas of China”
CHGIS raw input tables
CHGIS spatial data (boundaries)
CHGIS spatial data (points)
CHGIS spatial data (overlay)
points within boundaries
thiessen polygons
1820 changping quota (normalized by area)
CHGIS raw input tables
what are the temporal variables of “places?” placename - “Shawmut” to “Boston” feature types – dirt road to paved road location - point 1 to point 2 administrative status - township to county seat
independently changing variables place being tracked over time
tracking temporal “instances”
place objects – temporal instances each place object has many “instances” of change over time:
placenames each placename record may have many sub-types, such as: character sets, transliterations, or pronunciations character setstransliterations
feature classification schema Alexandria Digital Library (ADL) – Feature Type Thesaurus Bureau of Land Management, Taipei – NGIS cartographic elements hydrographic features. seas.. oceans... ocean currents
feature types - hierarchical model hydrographic features -aquifers -drainage basins -streams --rivers ---rapids physiographic features -cliffs Alexandria Digital Library (ADL) Feature Type Thesaurus:
feature types – relational model each feature instance may have many “feature types” feature types may have many associated feature types and glosses associated feature types glosses
feature types change over time a private residence is donated to a temple, affecting tax revenues both for the building and the fields under cultivation at different times the same place object may have different temporal instances of a particular feature type
temporal instances to track changing features a single building may have many functions over time, recorded with multiple “instances” of feature types:
feature types – changes over time tracking a building’s changing “feature types” associated feature types glosses
integrating feature type thesauri
cross-searching feature type thesauri list of available Feature Type thesauri: 1 – Alexandria Digital Library Feature Types (1156 terms) 2 – River Research Group Feature Types (435 terms) 3 – Taiwan NGIS Feature Types (2000 terms) 4 – CHGIS (220 terms)
rough classification for interoperability each system classifies its contents NIMA Classes: A – Admin Area H = Hydro L = Land Area P = Pop Place R = Road S = Spot T = Contour U = Undersea V = Vegetation
searching by rough class hydrographic features search
comparing contents among thesauri list of available “hydrographic” Feature Types in each thesauri: 1 – Alexandria Digital Library Feature Types (158 / 1156 terms) 2 – River Research Group Feature Types (380 / 435 terms) 3 – Taiwan NGIS Feature Types (200 / 2000 terms) 4 – CHGIS (37 /220 terms)
evolving interoperability
concordance table Records in the unrelated thesauri Concordance table between thesauri
proposed actions 1.use ADL Feature Types as a basic standard for interoperable searching and Unicode (UTF-8) 2.specialized thesauri submitted to the testbed should include a linktable with: - Feature Type definition - Feature Type Enlgish language gloss - Feature Type match to ADL type - “locale” for the Feature Type - date range for the Feature Type 4.Testbed editor enlists help of submitters to identify concordances between their own Feature Types and those in other thesauri 3.testbed editor identifies ADL types with high numbers of submitted specialized sub-types
proposed actions 1.use ADL Feature Types as a basic standard for interoperable searching and Unicode (UTF-8) 2.specialized thesauri submitted to the testbed should include a linktable with: - Feature Type definition - Feature Type Enlgish language gloss - Feature Type match to ADL type - “locale” for the Feature Type - date range for the Feature Type 4.Testbed editor enlists help of submitters to identify concordances between their own Feature Types and those in other thesauri 3.testbed editor identifies ADL types with high numbers of submitted specialized sub-types date formats 1) yyyy-mm-ddThh:mm:ss.sss or T09:40: Where T = Time 2) Periodic events are indicated with a P then a value and unit measure, as in: yyyy-mmPunitY or / /P1M 3) BCE dates take the letter “B” as a prefix, so B221 = BCE221. 4) Geologic dates in the distant past take the letters “K, M, G” as prefixes, meaning thousands, millions, and billions of years before the present, respectively. K16 = 16,000 years ago M120 = 120,000,000 years ago G3 = 3,000,000,000 years ago ISO – 8601 Date and Time Notation “with extension for older dates” Open GIS draft suggestion:
CHGIS: China Historical GIS Project website: fas.harvard.edu/~chgis