Announcing usbills.ai

We’re excited to announce the launch of usbills.ai, an open source project that aims to help the general public understand US federal legislation through automated analysis.

The Project

The Federal Bill Statistics project combines traditional natural language processing with large language models to provide accessible analysis of US legislation.

For each bill that passes through Congress, we provide:

• Plain language summaries and ELI5 (Explain Like I’m 5) explanations
• Key issue identification and topic analysis
• Readability metrics and complexity statistics
• Detailed section-level analysis, including focus on financial amounts
• Daily updates from the official govinfo.gov API

Data Collection and Sources

We retrieve bills directly from the govinfo.gov API, operated by the Government Publishing Office (GPO). For each bill:

• We download the official metadata and XML version
• Parse and extract the bill content, structure, and key information
• Store this in a standardized format for analysis
• Maintain links to original source documents for verification

Analysis Process

Traditional NLP Analysis

• Tokenization (breaking text into individual words)
• Syntactic parsing and POS tagging
• Named entity recognition for organizations, people, locations
• Basic statistics like word counts and sentence lengths

LLM-Based Analysis

• Plain language summaries of bill sections
• Key issue identification
• ELI5 descriptions
• Commentary on potential implications

Key Metrics

The platform provides several metrics to help understand bills:

Basic Statistics

• Section counts
• Word/token counts
• Character counts
• Sentence counts

Language Statistics

• Parts of speech distribution
• Key phrase frequency
• Entity mentions (organizations, people, places)

Complexity Metrics

• Average word length
• Average sentence length
• Token entropy (vocabulary diversity)
• Automated Readability Index (ARI)

Data Access

All data and code is freely available:

• Browse bills at usbills.ai
• Download complete JSON data from usbills.ai/index.json
• View source code at github.com/alea-institute/federal-bill-statistics

Important Notes

AI Content Limitations

Content marked with an AI tag:

• May contain errors or misinterpretations
• Should not be treated as legal advice
• Is meant to supplement, not replace, human analysis
• May reflect biases present in training data

Reading Original Sources

We strongly encourage users to:

• Read the original bill text using provided govinfo.gov links
• Verify information against official sources
• Consult qualified legal professionals for legal advice
• Use our analysis as a starting point for deeper research

Contributing

We welcome contributions from the community! Visit our GitHub repository to:

• Submit pull requests
• Report issues and bugs
• Suggest new features
• Help improve documentation

Looking Forward

This is just the beginning for the Federal Bill Statistics project. We plan to continue expanding the analysis capabilities while maintaining our commitment to openness and transparency.

Visit usbills.ai to explore the platform and learn more.

Please reach out to us at hello@aleainstitute.ai with any questions or feedback.