Classification in Twitter via Compressive Sensing

Dimitrios Milioris (1, 2, 3, 4)

1 EVA - Wireless Networking for Evolving & Adaptive Applications
2 X - École polytechnique
3 LINCS - Laboratory of Information, Network and Communication Sciences
4 Alcatel-Lucent Bell Labs France [Nozay]

In this paper we introduce a novel low dimensional method to perform topic detection and classification in Twitter. The proposed method first employs Joint Complexity to perform topic detection. Then, based on the nature of the data, we apply the theory of Compressive Sensing to perform topic classification by recovering an indicator vector, while reducing significantly the amount of information from tweets. In this paper we exploit datasets in various languages collected by using the Twitter streaming API, and achieve increased classification accuracy when comparing to state-of-the-art methods based on bag-of- words, along with several reconstruction techniques.

Domaines: Traitement du texte et du document; Traitement du signal et de l'image [eess.SP]

Dimitrios Milioris. Classification in Twitter via Compressive Sensing. IEEE International Conference on Computer Communications (INFOCOM), Apr 2015, Hong Kong, Hong Kong SAR China. ⟨hal-01138337⟩