Opinion Learning without Emotional Words

This paper shows that a detailed, although non-emotional, description of event or an action can be a reliable source for learning opinions. Empirical results show the practical utility of our approach and its competitiveness in comparison with previously used methods. 1 Motivation Humans can infer opinion from details of event or action description: I saw it several times can signal a positive opinion in a movie review; there was a long wait may indicate a negative opinion about health care services or a positive opinion about entertainment. We show that, under certain conditions, quantitative (few, small) and stance (probably) indicators extracted from texts provide for successful machine learning. Opinion learning has mainly studied polarity of texts (I enjoyed this movie expresses positive polarity, I hated the film – negative polarity). With an increased supply of free-form, unstructured or loosely structured texts, learning opinions can benefit from an assessment of the parame...

• Type: Conference
• Year: 2009
• Where: AI
• Authors: Marina Sokolova, Guy Lapalme
• Keywords: AI 2009, Artificial Intelligence, Descriptive Words, Opinion Learning, Positive Opinion