Sarcasm is a linguistic expression often used to communicate the opposite of what is said, usually something that is very unpleasant with an intention to insult or ridicule. Inherent ambiguity in sarcastic expressions, make sarcasm detection very difficult. In this work, we focus on detecting sarcasm in textual conversations from various social networking platforms and online media. To this end, we develop an interpretable deep learning model using multi-head self-attention and gated recurrent units. Multi-head self-attention module aids in identifying crucial sarcastic cue-words from the input, and the recurrent units learn long-range dependencies between these cue-words to better classify the input text. We show the effectiveness of our approach by achieving state-of-the-art results on multiple datasets from social networking platforms and online media. Models trained using our proposed approach are easily interpretable and enable identifying sarcastic cues in the input text which contribute to the final classification score. We visualize the learned attention weights on few sample input texts to showcase the effectiveness and interpretability of our model.

Polarization in online social networks has gathered a significant amount of attention in the research community and in the public sphere due to stark disagreements with millions of participants in topics surrounding politics, climate, the economy and other areas where an agreement is required. There are multiple approaches to investigating the scenarios in which polarization occurs and given that polarization is not a new phenomenon but that its virality may be supported by the low cost and latency messaging offered by online social media platforms; an investigation into the intrinsic dynamics of online opinion evolution is presented for complete networks. Extending a model which utilizes the Binary Voter Model (BVM) to examine the effect of the degree of freedom for selecting contacts based upon homophily, simulations show that different opinions are reinforced for a period of time when users have a greater range of choice for association. The facility of discussion threads and groups formed upon common views further delays the rate in which a consensus can form between all members of the network. This can temporarily incubate members from interacting with those who can present an alternative opinion where a voter model would then proceed to produce a homogeneous opinion based upon pairwise interactions.