Research Area:  Machine Learning
Nowadays, people often use social media platforms to express their feelings and opinions of a certain event in a very informal language. Complex ways of expressing opinions by different people make it difficult to determine the actual sentiments. Different elements that influence these sentiments are briefly discussed in this paper. To do this, only the content of that tweet is not enough; there is an emerging need to find some generic approach for sarcasm detection on Twitter. Our proposed framework concentrate not only on insights of tweets but also shades light on important factor of user behavior and its influence on other users. This factor is studied enormously in the field of psychology, but very few have worked on this in the domain of recognizing sarcasm in texts. This suggested method also considers context-based evaluation, based on data acquired from past experience. Context plays important role in determining user behavior, and it should not be ignored while detecting mockery. Our framework proposes to record user behavior pattern and personality traits along with context information. Accessing this information along with existing sarcasm-detection mechanism would help us to achieve generic approach to detect sarcasm on Twitter.
Keywords:  
Sarcasm Detection
Twitter
Machine Learning
Deep Learning
social media
Author(s) Name:  Nitin Malave, Sudhir N. DhageEmail author
Journal name:  
Conferrence name:  Intelligent Systems, Technologies and Applications
Publisher name:  Springer
DOI:  https://doi.org/10.1007/978-981-13-6095-4_5
Volume Information:  
Paper Link:   https://link.springer.com/chapter/10.1007/978-981-13-6095-4_5