AraBERT and Farasa Segmentation Based Approach For Sarcasm and Sentiment Detection in Arabic Tweets
This work addresses sarcasm and sentiment detection for Arabic social media users, but it is incremental as it applies existing methods to a specific dataset and task.
The paper tackled sarcasm and sentiment detection in Arabic tweets by preprocessing the ArSarcasm-v2 dataset and experimenting with transformer models like AraELECTRA and AraBERT, achieving seventh place in sarcasm detection and fourth in sentiment detection in the EACL WANLP-2021 Shared Task 2.
This paper presents our strategy to tackle the EACL WANLP-2021 Shared Task 2: Sarcasm and Sentiment Detection. One of the subtasks aims at developing a system that identifies whether a given Arabic tweet is sarcastic in nature or not, while the other aims to identify the sentiment of the Arabic tweet. We approach the task in two steps. The first step involves pre processing the provided ArSarcasm-v2 dataset by performing insertions, deletions and segmentation operations on various parts of the text. The second step involves experimenting with multiple variants of two transformer based models, AraELECTRA and AraBERT. Our final approach was ranked seventh and fourth in the Sarcasm and Sentiment Detection subtasks respectively.