update readme

README.md

@@ -1,20 +1,58 @@
+---
+license: agpl-3.0
+task_categories:
+- text-classification
+- code-understanding
+- semantic-analysis
+language:
+- en
+tags:
+- code
+- art
+- biology
+- synthetic
+- rust
+- ast
+- emoji
+- code-analysis
+pretty_name: rust_ast_emoji
+size_categories:
+- 100K<n<1M
+---
+
 # Rust AST Emoji Dataset
 
-This dataset contains Rust codebase AST (Abstract Syntax Tree) analysis with emoji mapping for code understanding and visualization.
+## Dataset Description
 
-## Dataset Structure
+- **Repository:** [GitHub Repository](https://github.com/your-repo/solfunmeme-dioxus)
+- **Paper:** [If applicable]
+- **Point of Contact:** [Your contact information]
+- **Huggingface Hub:** [Dataset link when published]
+
+### Dataset Summary
+
+This dataset contains Rust codebase AST (Abstract Syntax Tree) analysis with emoji mapping for code understanding and visualization. The dataset provides a unique perspective on code structure by mapping AST node types and extracted words to emojis, enabling creative code analysis and visualization.
+
+### Supported Tasks and Leaderboards
+
+- **Code Understanding:** Analyze code structure through emoji patterns
+- **Code Classification:** Identify code domains (Crypto, Web, i18n, etc.) through emoji signatures
+- **Code Visualization:** Create emoji-based code summaries and visualizations
+- **Pattern Recognition:** Discover common coding patterns through emoji frequency analysis
 
-- **Total Examples**: 15081
-- **Total Chunks**: 19
-- **Max File Size**: 10KB per chunk
-- **Max Files per Directory**: 10,000
+### Languages
 
-## Features
+The dataset contains Rust source code with English comments and identifiers.
 
-- `file_path`: Path to the original Rust source file
-- `timestamp`: Unix timestamp of analysis
-- `ast`: Full AST representation in JSON
-- `summary`: Analysis summary including:
+## Dataset Structure
+
+### Data Instances
+
+Each instance contains:
+- **file_path:** Path to the original Rust source file
+- **timestamp:** Unix timestamp of analysis
+- **ast:** Full AST representation in JSON format
+- **summary:** Analysis summary including:
   - `top_level_nodes`: Number of top-level AST nodes
   - `total_nodes`: Total number of AST nodes
   - `type_counts`: Count of each AST node type
@@ -23,14 +61,136 @@ This dataset contains Rust codebase AST (Abstract Syntax Tree) analysis with emo
   - `word_emoji_counts`: Emoji mapping for words
   - `emoji_counts_in_strings`: Emojis found in string literals
 
-## Usage
+### Data Fields
+
+- `file_path` (string): Path to the original Rust source file
+- `timestamp` (int64): Unix timestamp of analysis
+- `ast` (string): Full AST representation in JSON
+- `summary` (map): Analysis summary with nested fields:
+  - `top_level_nodes` (int64): Number of top-level AST nodes
+  - `total_nodes` (int64): Total number of AST nodes
+  - `type_counts` (map): Count of each AST node type
+  - `string_literals` (sequence): Extracted string literals
+  - `word_counts` (map): Word frequency analysis
+  - `word_emoji_counts` (map): Emoji mapping for words
+  - `emoji_counts_in_strings` (map): Emojis found in string literals
+
+### Data Splits
+
+- **train:** All analyzed Rust files
+
+## Dataset Creation
+
+### Source Data
+
+#### Initial Data Collection and Normalization
+
+The dataset was created by analyzing Rust source files from the solfunmeme-dioxus project, which includes:
+- Core application code
+- Vendor dependencies
+- Generated code
+- Test files
+
+#### Who are the source language producers?
+
+The source code was written by developers working on the solfunmeme-dioxus project, including contributions from the open-source community.
+
+### Annotations
+
+#### Annotation process
+
+The annotation process involved:
+1. **AST Parsing:** Using syn crate to parse Rust source files into ASTs
+2. **Emoji Mapping:** Mapping AST node types and extracted words to emojis based on semantic categories
+3. **Analysis:** Extracting string literals, word frequencies, and emoji patterns
+4. **Chunking:** Splitting large datasets into manageable chunks (1MB each)
+
+#### Who are the annotators?
+
+The annotations were generated automatically using a custom Rust script that implements emoji mapping based on predefined categories.
+
+### Personal and Sensitive Information
+
+The dataset contains only code analysis data and does not include personal or sensitive information. All file paths are relative to the project structure.
+
+## Additional Information
+
+### Dataset Curators
+
+The dataset was curated as part of the solfunmeme-dioxus project development process.
+
+### Licensing Information
+
+This dataset is licensed under AGPL-3.0, the same license as the source codebase.
+
+### Citation Information
+
+```bibtex
+@dataset{rust_ast_emoji,
+  title={Rust AST Emoji Dataset},
+  author={solfunmeme-dioxus contributors},
+  year={2024},
+  url={https://github.com/your-repo/solfunmeme-dioxus}
+}
+```
+
+### Contributions
+
+Contributions to improve the dataset, emoji mappings, or analysis methods are welcome through the project's GitHub repository.
+
+## Usage Examples
+
+### Basic Usage
+
+```python
+from datasets import load_dataset
+
+# Load the dataset
+dataset = load_dataset("your-username/rust_ast_emoji")
+
+# Access a sample
+sample = dataset["train"][0]
+print(f"File: {sample['file_path']}")
+print(f"Top-level nodes: {sample['summary']['top_level_nodes']}")
+print(f"Total nodes: {sample['summary']['total_nodes']}")
+```
+
+### Emoji Analysis
+
+```python
+# Analyze emoji patterns
+emoji_counts = sample['summary']['word_emoji_counts']
+for emoji, count in emoji_counts.items():
+    print(f"{emoji}: {count}")
+```
+
+### Code Domain Detection
+
+The dataset enables detection of code domains through emoji patterns:
+- 🌵 (Agave): Solana/blockchain code
+- 🎨 (CSS): Frontend/styling code  
+- 🔒 (Crypto): Security/cryptography code
+- 🌐 (i18n): Internationalization code
+
+## Technical Details
+
+### Chunking Strategy
+
+The dataset is split into chunks of maximum 1MB each to comply with Hugging Face and GitHub file size limits. Each chunk contains multiple code analysis examples.
+
+### Emoji Mapping Categories
 
-This dataset can be used for:
-- Code understanding and visualization
-- AST pattern analysis
-- Emoji-based code summarization
-- Codebase domain detection (Crypto, Web, i18n, etc.)
+The emoji mapping covers several categories:
+- **Rust Core:** Basic Rust language constructs (🦀⚙️, 🏛️🧱, etc.)
+- **Web/CSS:** Frontend and styling concepts (📏, 🧭, etc.)
+- **Crypto/Security:** Cryptography and security (🔒, 🔑, etc.)
+- **Project-Specific:** Domain-specific terms (🌵, 🌞, etc.)
+- **Internationalization:** i18n and localization (🌐, 🌍, etc.)
+- **Testing/Benchmarking:** Testing and performance (⏱️, 🏋️, etc.)
 
-## License
+### Performance Considerations
 
-AGPL-3.0 License
+The dataset is optimized for:
+- **Memory efficiency:** Compact JSON serialization
+- **Accessibility:** Small chunk sizes for easy loading
+- **Scalability:** Organized directory structure for large datasets