Chat Log Parser and Summarizer

A Python tool to analyze and summarize chat logs from .txt files using NLP techniques. Extracts key topics, message statistics, and generates summaries.

Features

• Parse single or multiple chat log files
• Extract speaker-specific messages (User/AI)
• Identify main topics using TF-IDF and lemmatization
• Generate summary statistics (message counts, keywords)

Prerequisites

• Python 3.12.4
• pip 24.0+

Setup

1. Create and activate virtual environment

python -m venv venv

# Windows:
venv\Scripts\activate

# Mac/Linux:
source venv/bin/activate

2. Install dependencies

pip install -r requirements.txt
python -m nltk.downloader stopwords wordnet punkt

Usage

1. For Single Chat File

python ai_chat_summarize_for_single_txt_file.py

Output Example:

Total messages: 4
User messages: 2
AI messages: 2

 Summary
 - The conversation had 15 exchanges
 - The user asked mainly about python and use
 - Most common keywords: python, use, hi, tell, sure

2. For Multiple Chat Files

python ai_chat_summarize_to_parse_all_txt_and_analysis.py

Output Example:

Total messages: 8
User messages: 4
AI messages: 4

Summary
The conversation had 26 exchanges
The user asked mainly about python and ai
Most common keywords: python, ai, data, hi, learn

3. Jupyter Notebook Option

jupyter notebook AI_Chat_Log_Summarizer_multiple_txt_parse.ipynb

Adding Screenshots

1. Create an assets/ folder:

   mkdir assets

2. Save screenshot (e.g., sample_output.png) in this folder

Project Structure

.
├── chat_log/                  # Folder for input chat logs (.txt files)
├── venv/                      # Virtual environment (ignored)
├── assets/                    # For screenshots and images
├── .gitignore
├── requirements.txt
├── README.md
├── ai_chat_summarize_for_single_txt_file.py
├── ai_chat_summarize_to_parse_all_txt_and_analysis.py
└── AI_Chat_Log_Summarizer_multiple_txt_parse.ipynb

Technical Details

• Uses NLTK for tokenization and lemmatization
• TF-IDF vectorization for keyword extraction
• Regular expression pattern matching for message parsing:

  PATTERN = r'(User|AI):\s*(.*?)(?=\n*User:|\n*AI:|\$)'

Troubleshooting

• If you get NLTK errors, re-run:

  python
  >>> import nltk
  >>> nltk.download('stopwords') 
  and so on (necessary libraries)

• For virtual environment issues:

  deactivate