CAP_coded_US_Congressional_bills
| Property | Value |
|---|---|
| Author | z-dickson |
| Framework | TensorFlow 2.8.2 |
| Training Accuracy | 92.68% |
| Validation Accuracy | 91.61% |
| Model URL | HuggingFace |
What is CAP_coded_US_Congressional_bills?
This is a specialized machine learning model designed to automatically categorize US Congressional bills according to the Comparative Agenda Project (CAP) coding scheme. Trained on approximately 250,000 Congressional bills spanning from 1950 to 2015, this model demonstrates robust performance in classifying legislative documents into standardized policy areas.
Implementation Details
The model is implemented using TensorFlow 2.8.2 and Transformers 4.19.3, utilizing the Adam optimizer with a learning rate of 5e-05. It maintains case sensitivity and achieves impressive accuracy metrics with a training loss of 0.1318 and validation loss of 0.2439.
- Training Precision: float32
- Extensive dataset coverage (65 years of legislative history)
- Implements CAP standardized coding taxonomy
- Case-sensitive text processing
Core Capabilities
- Automatic classification of congressional bills into policy areas
- High accuracy in both training (92.68%) and validation (91.61%) scenarios
- Standardized categorization following CAP methodology
- Processing of case-sensitive legislative text
Frequently Asked Questions
Q: What makes this model unique?
This model stands out for its specialized focus on US Congressional bills and its implementation of the Comparative Agenda Project's standardized coding scheme, making it particularly valuable for legislative analysis and political science research.
Q: What are the recommended use cases?
The model is ideal for political scientists, policy researchers, and legislative analysts who need to automatically categorize large volumes of congressional bills into standardized policy areas. It can be particularly useful for historical analysis of legislative trends and policy focus areas.
