English | 简体中文

GCA Analyzer

Introduction

GCA Analyzer is a Python package for analyzing group conversation dynamics using NLP techniques and quantitative metrics. It provides comprehensive tools for understanding participation patterns, interaction dynamics, content novelty, and communication density in group conversations.

Features

• Multi-language Support: Built-in support for Chinese and other languages through LLM models
• Comprehensive Metrics: Analyzes group interactions through multiple dimensions
• Automated Analysis: Finds optimal analysis windows and generates detailed statistics
• Flexible Configuration: Customizable parameters for different analysis needs
• Easy Integration: Command-line interface and Python API support

Tip

👁 Watch this repo so that you can be notified whenever there are fixes & updates.
📰 2025-01-22 GCA Analyzer v0.4.4 (beta) has been released!
📰 2025-01-14 GCA Analyzer v0.4.3 (beta) has been released!

Note

📝 If you have any questions or suggestions, please open an issue or contact etShaw-zh.
📝 You can find more information about GCA Analyzer in our documentation.

Quick Start

Installation

# Install from PyPI
pip install gca-analyzer

# For development
git clone https://github.com/etShaw-zh/gca_analyzer.git
cd gca_analyzer
pip install -e .

Basic Usage

1. Prepare your conversation data in CSV format with required columns:

conversation_id,person_id,time,text
1A,student1,0:08,Hello teacher!
1A,teacher,0:10,Hello everyone!

2. Run analysis:

python -m gca_analyzer --data your_data.csv

3. Descriptive statistics for GCA measures:

   The analyzer generates comprehensive statistics for the following measures:

   

   • Participation

     • Measures relative contribution frequency
     • Negative values indicate below-average participation
     • Positive values indicate above-average participation

   • Responsivity

     • Measures how well participants respond to others
     • Higher values indicate better response behavior

   • Internal Cohesion

     • Measures consistency in individual contributions
     • Higher values indicate more coherent messaging

   • Social Impact

     • Measures influence on group discussion
     • Higher values indicate stronger impact on others

   • Newness

     • Measures introduction of new content
     • Higher values indicate more novel contributions

   • Communication Density

     • Measures information content per message
     • Higher values indicate more information-rich messages

   Results are saved as CSV files in the specified output directory.

4. Visualizations for GCA measures:

   The analyzer provides interactive and informative visualizations for the following measures:

   

   • Radar Plots: Compare measures across participants
   • Distribution Plots: Visualize measure distributions

   Results are saved as interactive HTML files in the specified output directory.

Citation

If you use this tool in your research, please cite:

@software{gca_analyzer,
  title = {GCA Analyzer: Group Conversation Analysis Tool},
  author = {Xiao, Jianjun},
  year = {2025},
  url = {https://github.com/etShaw-zh/gca_analyzer}
}