The Role of Conversation Context for Sarcasm Detection in Online Interactions
TL
TL;DR
This work investigates several types of Long Short-Term Memory networks that can model both the conversation context and the sarcastic response and shows that the conditional LSTM network (Rocktäschel et al. 2015) and L STM networks with sentence level attention on context and response outperform the LSTm model that reads only the response.
Abstract
Computational models for sarcasm detection have often relied on the content of utterances in isolation. However, speaker’s sarcastic intent is not always obvious without additional context. Focusing on social media discussions, we investigate two issues: (1) does modeling of conversation context help in sarcasm detection and (2) can we understand what part of conversation context triggered the sarcastic reply. To address the first issue, we investigate several types of Long Short-Term Memory (LSTM) networks that can model both the conversation context and the sarcastic response. We show that the conditional LSTM network (Rocktäschel et al. 2015) and LSTM networks with sentence level attention on context and response outperform the LSTM model that reads only the response. To address the second issue, we present a qualitative analysis of attention weights produced by the LSTM models with attention and discuss the results compared with human performance on the task.
References39 items
1
Identification of nonliteral language in social media: A case study on sarcasm
2
Are Word Embedding-based Features Useful for Sarcasm Detection?
3
Creating and Characterizing a Diverse Corpus of Sarcasm in Dialogue
4
Hierarchical Attention Networks for Document Classification
5