Sentiment Analysis of Food Recipe Comments

Abstract

Sentiment analysis of food recipe comments is to identify user comments about the food recipes to the positive or the negative comments. The proposed method is suitable for analysing comments or opinions about food recipes by counting the polarity words on the food domain. The benefit of this research is to help users to choose the preferred recipes from different food recipes on online food communities. To analyse food recipes, the comments of each recipe from members of the community will be collected and classified to neutral, positive or negative comments. All recipes’ comment messages are processed using text analytics and the generated polarity lexicon. Therefore, the user can gain the information to make a smart decision. The evaluation of the comment analysis shows that the accuracy of neutral and positive comment classification is about 90%. In addition, the accuracy of negative comment identification is more than 70%.