Data-Driven vs. Semantic-Technology-Driven Tag-Based Video Location Estimation
Title | Data-Driven vs. Semantic-Technology-Driven Tag-Based Video Location Estimation |
Publication Type | Conference Paper |
Year of Publication | 2011 |
Authors | Choi, J., & Friedland G. |
Page(s) | 243-246 |
Other Numbers | 3219 |
Abstract | The following article describes two approaches todetermining the geo-coordinates of the recording place of Flickrvideos based on textual metadata. The systems are tested on theMediaEval 2010 Placing Task evaluation data, which consists of5091 unfiltered test videos and metadata records. The first systemis a data-driven approach that uses a heuristics based on thespatial variance of tags. The second one extends this heuristicsby using semantic technologies, such as extended Wordnet and ageographical gazetteer. The performance peaks at being able to |
Acknowledgment | This work was partially supported by funding provided through the National Geospatial-Intelligence Agency University Research Initiatives program (NGA NURI #HM11582-10-1-0008). Any opinions, findings, and conclusions or recommendations expressed in this material are those of the authors or originators and do not necessarily reflect the views of NGA. |
URL | http://www.icsi.berkeley.edu/pubs/speech/datadriven11.pdf |
Bibliographic Notes | Proceedings of the Fifth IEEE International Conference on Semantic Computing (ICSC 2011), pp. 243-246, Palo Alto, California |
Abbreviated Authors | J. Choi and G. Friedland |
ICSI Research Group | Audio and Multimedia |
ICSI Publication Type | Article in conference proceedings |