Datasets:
Formats:
csv
Languages:
Arabic
Size:
1K - 10K
Tags:
opinion-summarization
arabic-nlp
summarization-evaluation
sentiment-analysis
llm-evaluation
benchmark
License:
| statistic,value | |
| restaurants,100.0 | |
| total_reviews,1944.0 | |
| reviews_per_restaurant_min,10.0 | |
| reviews_per_restaurant_max,27.0 | |
| reviews_per_restaurant_mean,19.44 | |
| golden_summaries,100.0 | |
| gold_opinion_units,686.0 | |
| gold_entity_units,249.0 | |
| system_summaries,300.0 | |
| system_units,3312.0 | |
| annotation_rows,300.0 | |
| annotated_scores_rows_x_3_dimensions,900.0 | |
| GPT-4o_mean_relevance,0.872 | |
| GPT-4o_mean_faithfulness,0.703 | |
| GPT-4o_pct_positive,74.0 | |
| GPT-4o_total_repetitions,28.0 | |
| GPT-4o_total_conflicts,17.0 | |
| Claude_mean_relevance,0.767 | |
| Claude_mean_faithfulness,0.748 | |
| Claude_pct_positive,92.0 | |
| Claude_total_repetitions,41.0 | |
| Claude_total_conflicts,0.0 | |
| JAIS_mean_relevance,0.594 | |
| JAIS_mean_faithfulness,0.522 | |
| JAIS_pct_positive,74.0 | |
| JAIS_total_repetitions,45.0 | |
| JAIS_total_conflicts,41.0 | |
| gold_pct_positive,77.0 | |