On Evaluating the Contribution of Text Normalisation Techniques to Sentiment Analysis on Informal Web 2.0 Texts

@article{Lpez2017OnET,
  title={On Evaluating the Contribution of Text Normalisation Techniques to Sentiment Analysis on Informal Web 2.0 Texts},
  author={Alejandro Mosquera L{\'o}pez and Yoan Guti{\'e}rrez-V{\'a}zquez and Paloma Moreda},
  journal={Procesamiento del Lenguaje Natural},
  year={2017},
  volume={58},
  pages={29-36}
}
The writing style used in social media usually contains informal elements that can lower the performance of Natural Language Processing applications. For this reason, text normalisation techniques have drawn a lot of attention recently when dealing with informal content. However, not all the texts present the same level of informality and may not require additional pre-processing steps. Therefore, in this paper we explore the results of applying lexical normalisation applied to a sentiment… CONTINUE READING

5 Figures & Tables

Extracted Numerical Results

  • Therefore, in this paper we explore the results of applying lexical normalisation applied to a sentiment analysis classification task on Web 2.0 texts, shows more than a 2.6 % improvement over average F1 for the most informal data.
  • muestran una mejora de mas del 2.6 % sobre el F1 para los textos mas informales.
  • The results on Table 1 show how Sanders and Emotiblog texts obtained more than a 4 % and 3.5 % F1 improvement respectively on polarity classification by using the WN-Domain approach. All the F1 scores obtained during the experiments have been checked for statistical significance at 0.95 % confidence level.
  • Our experiments with Web 2.0 English texts consistently show higher average F1 over the original data.

Topics