Update README.md

README.md

@@ -1,40 +1,79 @@
----
-license: mit
-configs:
-- config_name: default
-  data_files:
-  - split: train
-    path: data/train-*
-dataset_info:
-  features:
-  - name: thread_id
-    dtype: int64
-  - name: category_id
-    dtype: int64
-  - name: category_name
-    dtype: string
-  - name: Question
-    dtype: string
-  - name: Question_token_count
-    dtype: int64
-  - name: Question_text_length
-    dtype: int64
-  - name: Question_language_probability
-    dtype: float64
-  - name: Answer
-    dtype: string
-  - name: Answer_token_count
-    dtype: int64
-  - name: Answer_text_length
-    dtype: int64
-  - name: Answer_language_probability
-    dtype: float64
-  - name: id
-    dtype: int64
-  splits:
-  - name: train
-    num_bytes: 1847812
-    num_examples: 2639
-  download_size: 901939
-  dataset_size: 1847812
----
+Dataset Summary
+This dataset contains questions from Nigerians on a dedicated healthcare forum and responses provided exclusively by licensed and trained healthcare professionals. It reflects health concerns within the Nigerian context, incorporating English and local colloquialisms. All answers are reliable, as the platform restricted responses to verified healthcare professionals, ensuring the quality and credibility of the information. The dataset comprises a total of 336,681 GPT-2 tokens.
+
+Data Collection
+The data was curated from the activities on an online platform's open forum page that offered the Nigerian public free advice and answers. It includes:
+
+1,731 questions
+2,667 replies/comments
+The data was extracted from a MySQL dump after the platform was closed.
+
+Data Cleaning
+The cleaning process involved:
+
+Deduplication: Removing duplicate entries to ensure uniqueness.
+Manual Review: Verifying the accuracy and relevance of the data.
+Normalization: Standardizing data formats to match the quality of other relevant datasets.
+
+---
+
+configs:
+- config_name: default
+  data_files:
+  - split: train
+    path: data/train-*
+    
+Dataset Structure
+Data Fields
+  features:
+  - name: thread_id
+    dtype: int64
+  - name: category_id
+    dtype: int64
+  - name: category_name
+    dtype: string
+  - name: Question
+    dtype: string
+  - name: Question_token_count
+    dtype: int64
+  - name: Question_text_length
+    dtype: int64
+  - name: Question_language_probability
+    dtype: float64
+  - name: Answer
+    dtype: string
+  - name: Answer_token_count
+    dtype: int64
+  - name: Answer_text_length
+    dtype: int64
+  - name: Answer_language_probability
+    dtype: float64
+  - name: id
+    dtype: int64
+  splits:
+  - name: train
+    num_bytes: 1847812
+    num_examples: 2639
+  download_size: 901939
+  dataset_size: 1847812
+---
+
+
+Usage
+This dataset can be valuable for:
+
+Natural Language Processing (NLP): Training models on Nigerian English and local colloquialisms.
+Healthcare Analysis: Understanding prevalent health concerns in Nigeria.
+Chatbot Training: Developing AI assistants for healthcare support in Nigerian contexts.
+
+license: mit
+
+Citation
+If you use this dataset in your research, please cite it as:
+
+@dataset{nigerian_healthcare_forum_qa,
+  title={Nigerian Healthcare Forum Q\&A Dataset},
+  year={2023},
+  publisher={Your Organization},
+  note={Data retrieved from a closed online platform's MySQL dump}
+}