Contract Review Interface

This Streamlit application provides an interface for reviewing contracts against a set of guidelines. It allows users to evaluate contract clauses, provide reasons for compliance or non-compliance, and suggest improvements.

Features

• Display contract text with selectable paragraphs
• Review guidelines one by one
• Evaluate guideline quality and suggest improvements
• Determine if a guideline is followed, not followed, or not applicable
• Provide reasons for guideline status with quality evaluation
• Add comments with quality evaluation
• Suggest updated clause text with improvement options
• Save review results to Google Sheets
• Navigate through guidelines with progress tracking

Requirements

• Python 3.6+
• Streamlit
• pandas
• streamlit-gsheets

Installation

1. Clone this repository:

   git clone https://github.com/SpotDraft/DataAnnotation.git
   cd DataAnnotation
   

2. Install the required packages:

   pip install -r requirements.txt
   

3. Set up your Google Sheets connection:

   • Create a Google Cloud project and enable the Google Sheets API
   • Create a service account and download the JSON key
   • Share your Google Sheet with the service account email

4. Create a .streamlit/secrets.toml file with your Google Sheets credentials:

   spreadsheet = "YOUR_SPREADSHEET_ID"
   [connections.gsheets]
    private_key_id = "YOUR_PRIVATE_KEY_ID"
    private_key = "YOUR_PRIVATE_KEY"
    client_email = "YOUR_CLIENT_EMAIL"
    client_id = "YOUR_CLIENT_ID"
    type = "YOUR_TYPE"
    project_id = "YOUR_PROJECT_ID"
    auth_uri = "YOUR_AUTH_URI"
    token_uri = "YOUR_TOKEN_URI"
    auth_provider_x509_cert_url = "YOUR_AUTH_PROVIDER"
    client_x509_cert_url = "YOUR_CLIENT_X509"
    universe_domain = "YOUR_DOMAIN"
   
   

Usage

Run the Streamlit app:

streamlit run app.py

Navigate to the provided URL to access the Contract Review Interface.

Data Structure

The application expects the following worksheets in your Google Sheet:

1. "Data": Contains the guidelines and contract text
2. "ReviewResults": Stores the review results

Input Data Format

The "Data" worksheet should have the following columns:

• id: Unique identifier for each guideline
• guideline: The text of the guideline
• contract: The full text of the contract (this will be the same for all rows)
• reason: ...
• status: ...
• comment: ...
• updated_clause_text: ...

Example:

id	guideline	contract	reason	status	comment	updated_clause_text
1	Ensure all parties are clearly identified	Full contract text here...	counterparty is missing	NOT_FOLLOWED	Please add counterparty	Counterparty is..
...	...	...

Output Data Format

The "ReviewResults" worksheet will store the review results with the following columns:

• id: Unique identifier for the guideline (matching the input data)
• guideline: The text of the guideline
• guideline_quality: Quality assessment of the guideline (Pending, Excellent, Good, Better, Bad)
• guideline_improvement: Suggested improvement area for the guideline
• guideline_improvement_other: Custom improvement suggestion if "Other" was selected
• status: Whether the guideline is FOLLOWED, NOT_FOLLOWED, or NOT_APPLICABLE
• reason: Explanation for the status
• reason_quality: Quality assessment of the reason (Pending, Excellent, Good, Better, Bad)
• reason_improvement_other: Custom improvement suggestion for the reason if "Other" was selected
• comment: Additional comments
• comment_quality: Quality assessment of the comment (Pending, Excellent, Good, Better, Bad)
• comment_improvement_other: Custom improvement suggestion for the comment if "Other" was selected
• selected_sources: Comma-separated list of paragraph numbers from the contract text that are relevant to this guideline
• update_clause_text: Suggested updated text for the relevant clause
• update_clause_improvement: Assessment of the updated clause text (NA, Changes too extensive, Does not meet guideline, Content inaccuracies, Other)
• update_clause_improvement_other: Custom improvement suggestion for the updated clause if "Other" was selected

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.