Studies in health technology and informatics · 2013 · 12 citations · 0 references
ComputingData ScienceHealth BehaviorData MiningMachine LearningNatural Language ProcessingOpinion MiningSentiment AnalysisSocial MediaSocial Network MiningSupport Vector MachinesText MiningAccuracyPolarity DetectionSentiment ClassificationSentiment FeaturesSocial Media SentimentSupport Vector ClassificationTextual Sentiment AnalysisTweet Sentiment Analysis
Online social networks play a vital role in daily life to share the opinions or behaviors on different topics. The data of social networks can be used to understand health-related behaviors. In this work, we used Twitter status updates to survey of smoking behaviors among the users. We introduce approach to classify the sentiment of smoke-related tweets into positive and negative tweets. The classifier is based on the Support Vector Machines (SVMs) and can achieve high accuracy up to 86%.