Sentiment analysis using a novel human computation game

  • Authors:
  • Claudiu-Cristian Musat Thisone;Alireza Ghasemi;Boi Faltings

  • Affiliations:
  • Artificial Intelligence Laboratory (LIA), Lausanne, Switzerland;Artificial Intelligence Laboratory (LIA), Lausanne, Switzerland;Artificial Intelligence Laboratory (LIA), Lausanne, Switzerland

  • Venue:
  • Proceedings of the 3rd Workshop on the People's Web Meets NLP: Collaboratively Constructed Semantic Resources and their Applications to NLP
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

In this paper, we propose a novel human computation game for sentiment analysis. Our game aims at annotating sentiments of a collection of text documents and simultaneously constructing a highly discriminative lexicon of positive and negative phrases. Human computation games have been widely used in recent years to acquire human knowledge and use it to solve problems which are infeasible to solve by machine intelligence. We package the problems of lexicon construction and sentiment detection as a single human computation game. We compare the results obtained by the game with that of other well-known sentiment detection approaches. Obtained results are promising and show improvements over traditional approaches.