In this paper we present our approach for the Social Event Detection Task 1 of the MediaEval 2013. We address the problem of event detection and clustering by learning a distance measure between two images in a supervised way. Then, we apply a variant of the Quality Threshold clustering to detect events and assign the images accordingly. We can show that the performance measures do not decrease for an increasing number of documents and report the results achieved for the challenge. 1. INTRODUCTION This paper presents our approach to tackle Task 1 of the MediaEval Social Event Detection 2013 Challenge [7]. The task is to cluster images into an unknown number of events in such a way that they belong to each other. For the required run only meta information like title and description may be used whereas for the general runs more information can be considered. Here, we only discuss an approach for the required run.
[1]
Steffen Rendle,et al.
Factorization Machines
,
2010,
2010 IEEE International Conference on Data Mining.
[2]
Yiannis Kompatsiaris,et al.
Social Event Detection at MediaEval 2012: Challenges, Dataset and Evaluation
,
2012,
MediaEval.
[3]
Laurie J. Heyer,et al.
Exploring expression data: identification and analysis of coexpressed genes.
,
1999,
Genome research.
[4]
Hila Becker,et al.
Learning similarity metrics for event identification in social media
,
2010,
WSDM '10.
[5]
Lars Schmidt-Thieme,et al.
Scalable Event-Based Clustering of Social Media Via Record Linkage Techniques
,
2011,
ICWSM.
[6]
Martin F. Porter,et al.
An algorithm for suffix stripping
,
1997,
Program.
[7]
Philipp Cimiano,et al.
Event-based classification of social media streams
,
2012,
ICMR.