As a platform for unsupervised data mining and pattern recognition, we use Correspondence Analysis on Twitter content from May to December 2015. The following data characteristics are well addressed: exponentially distributed data properties, and major imbalance between categories. Contextualization is supported. To both focus on informative resolution scale in one's data, and to handle large data sets, the granularity of point clouds offers benefits.
|Number of pages||19|
|Journal||CEUR Workshop Proceedings|
|Publication status||Published - 25 Jul 2016|
|Event||7th Conference and Labs of the Evaluation Forum: Information Access Evaluation meets Multilinguality, Multimodality and Visualization - University of Évora, Evora, Portugal|
Duration: 5 Sep 2016 → 8 Sep 2016
Conference number: 7
http://clef2016.clef-initiative.eu/ (Link to Event Website)