Upload README.md with huggingface_hub

README.md

@@ -1,17 +1,25 @@
 ---
-dataset_info:
-  features:
-  - name: Joke_Text
-    dtype: string
-  splits:
-  - name: train
-    num_bytes: 118901
-    num_examples: 510
-  download_size: 64151
-  dataset_size: 118901
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
+language:
+- ar
+license: mit
+task_categories:
+- text-generation
+- text-classification
+tags:
+- humor
+- arabic
+- jokes
+size_categories:
+- 1K<n<10K
 ---
+
+# Extracted Arabic Jokes Dataset
+
+## Dataset Description
+This dataset contains Arabic jokes extracted and cleaned from various PDF booklets using OCR and LLM-based post-processing. It was created to support research in Arabic Humor Generation.
+
+### Data Fields
+- `Joke_Text`: The full text of the joke in Arabic.
+
+### Source
+Extracted from public domain joke books.