Papers by Lamia Hadrich Belguith
Toward Qualitative Evaluation of Embeddings for Arabic Sentiment Analysis (2020.lrec-1)
Copied to clipboard
| Challenge: | Existing studies on Arabic sentiment analysis (SA) tasks focus on word embeddings to capture semantic and syntactic similarities, but Arabic language is characterized by its agglutination and morphological richness contributing to great sparsity. |
| Approach: | They propose several protocols to evaluate specific embeddings for Arabic sentiment analysis task. |
| Outcome: | The proposed embeddings are based on words and lemmas in Arabic sentiment analysis (SA) task. |