GeoNames extractors
GeoNames is a very large networked database of geographical locations and related data. GeoNames contains over eight million geographical names and 6.5 million unique features. GeoNames database is available for free of charge under a creative commons attribution license. GeoNames has a web service api with various different query types.
Wandora's GeoNames extractor performs GeoNames web service api requests and converts results to topic maps formats. Wandora is a graphical front end to GeoNames web service api, if you like. Extractor starts with menu option File > Extract > Other > GeoNames extractor.... Current extractor has several options as image below views. By default extractor performs free text search to GeoNames geographical database.
Available extractors are
| Tab | Description |
| Search | Performs free text search to GeoNames database. User may filter result set with optional continent, country, and feature class filters. All filters are selected using drop down selection. User can also require the search term is part of the place name. |
| Near by | Performs coordinate search to GeoNames database. User gives extractor latitude and longitude coordinate, radius in kilometers, and optional feature class. Result set contains geographical location in specified area. If user has already performed GeoNames extractions and has selected one geo-topic with latitude and longitude coordinates, user can get coordinates from the context topic. |
| Country info | Reads country information from GeoNames database. Country name is selected from drop down menu. Optionally user can write country name or multiple country names to the field. Special keyword ALL reads information for all known countries. If user has already performed GeoNames extractions or has country name available in topic for some other reason, user can pick up the country name from context topics. |
| Cities | Reads cities in given bounding box from GeoNames database. Bounding box has four different geographical coordinates:
If user has already performed GeoNames extractions and has suitable topic selected, Get context button reads available coordinates from the topic. For example, Country info extraction sets bounding box for each country topic, and user could use this bounding box to continue extracting cities of the country. |
| Children | Reads direct geo locations within given geo location known by GeoNames database. For example, if user gets children of Europe, the result set contains all countries in Europe. For example, if user gets children of France, the result set contains all first level administrative areas of France (but *not* cities as you might assume). Children extractor assumes you specify the parent geo location using special GeoNames identifier.
GeoNames identifier is a 7-digit number sequence. Whenever available, Wandora adds GeoNames originating topics subject identifier constructed with GeoNames identifier. For example country Algeria has GeoNames identifier 2589581. When Wandora creates a topic for Algeria, topic gets subject identifier http://www.geonames.org/2589581. One should note, the subject identifier actually resolves to a HTML page of Algeria hosted by GeoNames. Children extractor can read the GeoNames identifier from context topics, if they carry valid GeoNames subject identifier. Thus, Children extractor is usually *not* the first extractor used as it assumes user remembers identifier numbers. However, later on, when there is geo topics available, it is rather easy to use existing identifiers. |
| Hierarchy | Reads upward hierarchy of given geo location. Upward hierarchy is geo location chain from current geo location to the globe. For example, hierarchy of Finland contains geo locations Europe and Globe.
Hierarchy extractor requires GeoNames identifier to perform request. Read children line above for details about GeoNames identifiers. |
| Neighbours | Reads geo locations that are geographically neighbours of given geo location. For example, Finland's neighbours are Norway, Russia, and Sweden. Current web service returns neighbours for countries only.
Neighbours extractor requires GeoNames identifier to perform request. Read children line above for details about GeoNames identifiers. |
| Siblings | Reads geo locations that locate in same geo area and are same type than given geo location. For example, siblings of Finland are all other countries in Europe.
Siblings extractor requires GeoNames identifier to perform request. Read children line above for details about GeoNames identifiers. |
| Weather | Reads latest weather observation of nearest weather station at given geo coordinates. Weather observation contains for example
Weather observation request requires latitude and longitude near-by coordinates of the weather station. |
| Wikipedia search | GeoNames database contains also geographical locations provided by Wikipedia. Wikipedia search extractor performs free text search to Wikipedia data and returns geographical locations that have search word in their Wikipedia page. |
| Wikipedia b-box | Wikipedia b-box tab reads geographical locations within given bounding box provided by Wikipedia. |
Example
TODO
Known issues
- Performing many sequential extractions may cause identity doubling. Identity doubling means topic map contains multiple topics that are essentially identical and should be merged but for some reason topics don't share subject identifier, base name, or subject locator, and topics are never merged. For example, extracting country info and then continuing with neighbour extractions results duplicate topics for all neighbour countries. This issue will be fixed in near future releases!
- Two Wikipedia related GeoNames extractors generate topic map fragments are not compatible with other GeoNames extrators. You have to merge equivalent topics manually to prevent identity duplication.
