arXiv:2005.11424
Transformer-based Context-aware Sarcasm Detection in Conversation Threads from Social Media
원문 초록
We present a transformer-based sarcasm detection model that accounts for the context from the entire conversation thread for more robust predictions. Our model uses deep transformer layers to perform multi-head attentions among the target utterance and the relevant context in the thread. The context-aware models are evaluated on two datasets from social media, Twitter and Reddit, and show 3.1% and 7.0% improvements over their baselines. Our best models give the F1-scores of 79.0% and 75.0% for the Twitter and Reddit datasets respectively, becoming one of the highest performing systems among 36 participants in this shared task.
함께 살펴볼 논문
공통 태그를 근거로 연결했습니다. 실제 인용 관계나 논문의 우수성을 뜻하지 않습니다.
RCLMuFN: Relational Context Learning and Multiplex Fusion Network for Multimodal Sarcasm Detection
Gender Bias Mitigation for Bangla Classification Tasks
An Innovative CGL-MHA Model for Sarcasm Sentiment Recognition Using the MindSpore Framework
Sarcasm Detection in a Less-Resourced Language
InterCLIP-MEP: Interactive CLIP and Memory-Enhanced Predictor for Multi-modal Sarcasm Detection
CofiPara: A Coarse-to-fine Paradigm for Multimodal Sarcasm Target Identification with Large Multimodal Models
토론
이 논문과 연결된 포럼 글이 아직 없습니다.