Title :
Sentiment Regression: Using Real-Valued Scores to Summarize Overall Document Sentiment
Author :
Drake, Adam ; Ringger, Eric ; Ventura, Dan
Author_Institution :
Comput. Sci. Dept., Brigham Young Univ., Provo, UT
Abstract :
In this paper, we consider a sentiment regression problem: summarizing the overall sentiment of a review with a real-valued score. Empirical results on a set of labeled reviews show that real-valued sentiment modeling is feasible, as several algorithms improve upon baseline performance. We also analyze performance as the granularity of the classification problem moves from two-class (positive vs. negative) towards infinite-class (real-valued).
Keywords :
classification; learning (artificial intelligence); regression analysis; support vector machines; text analysis; SVM regression algorithm; learning algorithms; overall document sentiment summarization; real-valued score; real-valued sentiment modeling; sentiment regression problem; written text classification; Classification algorithms; Computer science; Filtering; Fires; Games; Labeling; Machine learning algorithms; Mutual information; Organizing; Performance analysis;
Conference_Titel :
Semantic Computing, 2008 IEEE International Conference on
Conference_Location :
Santa Clara, CA
Print_ISBN :
978-0-7695-3279-0
Electronic_ISBN :
978-0-7695-3279-0
DOI :
10.1109/ICSC.2008.67
