Sentiment Analysis gains in interest due to the large amount of potential applications and the increasing number of opinions expressed in particular in the Web. The focus of this paper is the development of a framework on top of sentiment analysis for detecting contradictions. First, we introduce a statistical model of contradictions based on a mean value and the variance of sentiments among different posts. It can be used to analyze and track sentiment evolution over time, to identify interesting trends and patterns or even to enable argument extraction. Using synthetic datasets, we demonstrate the effectiveness of our method in capturing contradictions on noisy data. Inspired by this model, which has proven to be effective and efficient for numeric sentiments, we are trying to generalize it for arbitrary opinion data and outline a universal framework which can be efficiently used on a large scale. We discuss various problems and challenges of such a formulation and outline the scope of our future work in this direction.

Towards a Framework for Detecting and Managing Opinion Contradictions

Tsytsarau, Mikalai;Palpanas, Themistoklis
2011-01-01

Abstract

Sentiment Analysis gains in interest due to the large amount of potential applications and the increasing number of opinions expressed in particular in the Web. The focus of this paper is the development of a framework on top of sentiment analysis for detecting contradictions. First, we introduce a statistical model of contradictions based on a mean value and the variance of sentiments among different posts. It can be used to analyze and track sentiment evolution over time, to identify interesting trends and patterns or even to enable argument extraction. Using synthetic datasets, we demonstrate the effectiveness of our method in capturing contradictions on noisy data. Inspired by this model, which has proven to be effective and efficient for numeric sentiments, we are trying to generalize it for arbitrary opinion data and outline a universal framework which can be efficiently used on a large scale. We discuss various problems and challenges of such a formulation and outline the scope of our future work in this direction.
2011
Proceedings of the IEEE International Conference on Data Mining
New York
IEEE Computer Society Press
9781467300056
Tsytsarau, Mikalai; Palpanas, Themistoklis
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11572/88862
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? ND
social impact