# Sarcasm Analysis using Conversation Context

arXiv: https://arxiv.org/abs/1808.07531 · HTML: https://mlsanity.com/paper/1808.07531/

## Key facts

- Cited by 17 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Aug 2018.

## Built on by (most cited first, 17 of 17)

- [When did you become so smart, oh wise one?! Sarcasm Explanation in Multi-modal Multi-party Dialogues](https://mlsanity.com/paper/2203.06419/) (2022)
- [SarcasmBench: Towards Evaluating Large Language Models on Sarcasm Understanding](https://mlsanity.com/paper/2408.11319/) (2024)
- ["Laughing at you or with you": The Role of Sarcasm in Shaping the Disagreement Space](https://mlsanity.com/paper/2101.10952/) (2021)
- [Generalizable Sarcasm Detection Is Just Around The Corner, Of Course!](https://mlsanity.com/paper/2404.06357/) (2024)
- [A Survey on Automated Sarcasm Detection on Twitter](https://mlsanity.com/paper/2202.02516/) (2022)
- [Analyzing Toxicity in Deep Conversations: A Reddit Case Study](https://mlsanity.com/paper/2404.07879/) (2024)
- [Computational Sarcasm Analysis on Social Media: A Systematic Review](https://mlsanity.com/paper/2209.06170/) (2022)
- [BNS-Net: A Dual-channel Sarcasm Detection Method Considering Behavior-level and Sentence-level Conflicts](https://mlsanity.com/paper/2309.03658/) (2023)
- [Dimensions of Online Conflict: Towards Modeling Agonism](https://mlsanity.com/paper/2311.03584/) (2023)
- [KoCoSa: Korean Context-aware Sarcasm Detection Dataset](https://mlsanity.com/paper/2402.14428/) (2024)
- ["When Words Fail, Emojis Prevail": Generating Sarcastic Utterances with Emoji Using Valence Reversal and Semantic Incongruity](https://mlsanity.com/paper/2305.04105/) (2023)
- [Evaluating Large Language Models Against Human Annotators in Latent Content Analysis: Sentiment, Political Leaning, Emotional Intensity, and Sarcasm](https://mlsanity.com/paper/2501.02532/) (2025)
- [Efficient Toxicity Detection in Gaming Chats: A Comparative Study of Embeddings, Fine-Tuned Transformers and LLMs](https://mlsanity.com/paper/2510.17924/) (2025)
- [Interpretable Multi-Head Self-Attention model for Sarcasm Detection in social media](https://mlsanity.com/paper/2101.05875/) (2021)
- [IISERB Brains at SemEval 2022 Task 6: A Deep-learning Framework to Identify Intended Sarcasm in English](https://mlsanity.com/paper/2203.02244/) (2022)
- [Sarcasm Detection on Reddit Using Classical Machine Learning and Feature Engineering](https://mlsanity.com/paper/2512.04396/) (2025)
- [A GAN and LLM-Driven Data Augmentation Framework for Dynamic Linguistic Pattern Modeling in Chinese Sarcasm Detection](https://mlsanity.com/paper/2604.08381/) (2026)

## Abstract (arXiv)

Computational models for sarcasm detection have often relied on the content of utterances in isolation. However, the speaker's sarcastic intent is not always apparent without additional context. Focusing on social media discussions, we investigate three issues: (1) does modeling conversation context help in sarcasm detection; (2) can we identify what part of conversation context triggered the sarcastic reply; and (3) given a sarcastic post that contains multiple sentences, can we identify the specific sentence that is sarcastic. 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 current turn. We show that LSTM networks with sentence-level attention on context and current turn, as well as the conditional LSTM network (Rocktaschel et al. 2016), outperform the LSTM model that reads only the current turn. As conversation context, we consider the prior turn, the succeeding turn or both. Our computational models are tested on two types of social media platforms: Twitter and discussion forums. We discuss several differences between these datasets ranging from their size to the nature of the gold-label annotations. To address the last two issues, 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 two tasks.