Abstract

Place information is important in building a wide variety of geographic applications, such as location-based services for mobile devices. Given these needs, techniques for automatically building place-name datasets are desperately needed. Traditionally, place datasets can be obtained from structured sources such as DBpedia, LinkedGeoData, Wikimapia, Google Places API, and OpenStreetMap. These structured sources provide static information and do not always incorporate the latest changes. Also, commercial sources such as the Google Places API maintain high-quality location data, but many restrictions apply to obtaining and using their data (eg, usage limits). In contrast, the volume of place information publicly available on the Web is very large and grows rapidly. In addition, the unstructured nature of webpages allows them to change frequently, and up-to-date information about places is often first available on the Web. The challenge is how to extract accurate and timely geographic information from the Web to build place-name datasets.

Date

January 1, 1970

Authors

Ying Zhang, Chaopeng Li, Liming Du, Shaowen Liu, Yao-Yi Chiang, Craig A Knoblock, Sanjay Singh

Conference

Proc. 19th AGILE Conf. Geograph. Inf. Sci.

Pages

1-6