Jeocrowd - Collaborative Searching of User-Generated Point Datasets George Lamprianidis Dieter Pfoser Institute for the Management of Information Institute for the Management of Information Systems Systems G. Mpakou 17 G. Mpakou 17 11526 Athens, Greece 11526 Athens, Greece
[email protected] [email protected] ABSTRACT tion ambiguity. Consider here the example of geotagged Flickr pho- Geospatial data has become an important resource in today’s Web tos. While many of them might refer to the same POI, the recorded applications not only as type of content, but also as metadata. De- coordinate information hardly matches up due to positioning inac- spite its undisputed usefulness, issues need to be addressed with re- curacy, or, simply, wrong geotagging. For example, for large scale spect to the availability, the accuracy, and the cost of the data. The objects different users will record different coordinates. Some of advent of Web2.0 created several creative-commons initiatives ad- them might even capture the POI from a distance, which will set the dressing geospatial dataset creation and countless (mobile) applica- coordinates completely off. tions have been producing large amounts of point cloud datasets. The objective of this work is to transform available UG geocon- In this work, we demonstrate how to query user-contributed point- tent into meaningful chunks of information, geospatial datasets, ob- cloud data using a collaborative Web-based approach. tained with simplicity and speed comparable to that of Web-based search. To achieve this task, we propose a search method utilizing crowdsourcing concepts implemented as a Web-based, collaborative Categories and Subject Descriptors search tool termed Jeocrowd1.