Upload README.md with huggingface_hub

README.md

@@ -2,43 +2,56 @@
 license: apache-2.0
 ---
 
-# 🇮🇩 Indonesian Regulation PDFs
+# 🇮🇩 Indonesian Regulation Parsed Dataset
 
-A high-volume, web-scraped collection of **250,000+ Indonesian legal documents in PDF**, aggregated from \~350,000 public URLs. This dataset enables **legal NLP, document analysis, and regulation-aware AI applications** in Bahasa Indonesia.
+This dataset contains the **parsed content of over 250,000 Indonesian regulations**, extracted directly from PDFs at the article level — **without chunking** — resulting in **3.5+ million rows** representing individual articles and clauses.
 
 ---
 
-## ⚡ Key Highlights
+## 📄 Dataset Overview
 
-* **Format**: Archived `.zip` files (each \~5,000 PDFs)
-* **Total Docs**: \~250K successfully downloaded
-* **Scraped From**: Government regulation portal
-* **Cloud Pipeline**: Scraped using **6 Google Colab nodes**, pushed directly to Hugging Face
-* **Duration**: \~200 hours distributed scraping in total
-* **Failures**: Some links unreachable or invalid
+* **Source**: Parsed from the original PDF collection of Indonesian regulations (\~250K documents)
+* **Parsed Units**: Articles, clauses, or sections as atomic units (no chunking applied)
+* **Size**: Over **3.5 million rows** of parsed text data
+* **Token Count**: Approximately **2 billion tokens** in total
 
 ---
 
-## 🧠 Ideal For
+## ⚙️ Processing Pipeline
 
-* Legal document retrieval & classification
-* Training LLMs on regulatory content
-* Building regulation-aware assistants
-* Multilingual or cross-domain NLP in Bahasa Indonesia
+* **Data Flow**:
+
+  1. Pull PDFs from the **[ID\_REG repository](https://huggingface.co/datasets/Azzindani/ID_REG)**
+  2. Parse and extract articles using automated PDF text extraction methods
+  3. Push cleaned, parsed text to this Hugging Face repository
+
+* **Infrastructure**:
+
+  * Utilized **6 simultaneous Google Colab instances** for parallel processing
+  * Cloud-to-cloud architecture enables direct transfer without local storage bottlenecks
+  * Total processing time: \~30 hours
+
+---
+
+## 🧠 Use Cases
+
+* Fine-grained legal text analysis and search
+* Article-level classification, summarization, or information retrieval
+* Training large language models on legal text with natural article boundaries
+* Developing Indonesian regulation compliance tools
 
 ---
 
 ## ⚠️ Disclaimer
 
-This dataset is intended **solely for research and development**. It contains **publicly accessible legal documents** collected via ethical scraping practices.
-No warranties are made regarding the accuracy, completeness, or legal validity of the content.
+This dataset is provided **exclusively for research and development** purposes. It contains parsed content derived from publicly available legal PDFs, and no guarantees are made regarding completeness or legal accuracy.
 
 ---
 
 ## 🙏 Acknowledgments
 
-* Dataset hosted on **[Hugging Face](https://huggingface.co/)** — thanks for providing an amazing platform for sharing open datasets.
-* Data scraping powered by **[Google Colab](https://colab.research.google.com/)** — enabling scalable cloud computing to speed up data collection.
+* Hosted on **[Hugging Face](https://huggingface.co/)** — a wonderful platform for open data sharing
+* Processing powered by **[Google Colab](https://colab.research.google.com/)** — scalable cloud resources that made fast parallel parsing possible
 
 ---