Abstract
Conference Title: 2017 International Conference on Engineering & MIS (ICEMIS) Conference Start Date: 2017, May 8 Conference End Date: 2017, May 10 Conference Location: Monastir, Tunisia Sarcasm is a special form of irony or satirical wit in which people convey the opposite of what they mean. Sarcasm largely increases in social networks, especially in Twitter. Detecting sarcasm in tweets improves the automatic analysis tools that analyze the data to provide or enhance customer service and fabricate or enhance a product. Also, there are few studies that focus on detecting Arabic sarcasm in tweets. Consequently, we propose a classifier model that detects Arabic-sarcasm tweets by classifying them as sarcastic by setting some features that may declare a tweet as sarcastic using Weka. We evaluated our model through recall, precision, and f-score measurements that gave 0.659, 0.710, and 0.676 values, respectively, which these results are high especially when it comes to Arabic.