InsightSummarizer is an AI-powered text summarization tool that condenses lengthy documents into concise, easy-to-digest summaries. Whether you're dealing with books, articles, or reports, InsightSummarizer helps you grasp the essential points quickly, saving you time and enhancing understanding.

• Chapter-wise Summarization: Generate summaries for individual chapters, focusing on specific sections of interest.
• Entire Document Summarization: Summarize the whole content when documents lack clear divisions.
• Advanced Natural Language Processing (NLP): Powered by cutting-edge NLP techniques to intelligently extract the most relevant information.
• User-Friendly Interface: An intuitive interface makes the summarization process seamless for users of all technical backgrounds.
• Email Notifications: Receive summarized content directly in your inbox for convenience.

InsightSummarizer leverages the T5-small pretrained model from HuggingFace Transformers to generate accurate and coherent summaries. The process involves:

1. Text Segmentation: The document is divided into chunks, either by chapters or as a whole.
2. Tokenization: Chunks are tokenized using T5Tokenizer for compatibility with the T5 model.
3. Summary Generation: The tokenized text is processed by the T5ForConditionalGeneration model to produce summary token IDs.
4. Decoding: Summary tokens are decoded back into human-readable text using the T5Tokenizer's decode() function.

• Python 3.x
• Required Python packages listed in requirements.txt

1. Clone the Repository

   git clone https://github.com/zjnbwxq/InsightSummarizer.git

2. Navigate to the Project Directory

   cd InsightSummarizer

3. Install Dependencies

   pip install -r requirements.txt

Run InsightSummarizer via CLI:

python3 summarizer_cli.py --path <path-to-your-document>

1. Configure Email Settings

   Update sender_address and sender_pass in mail.py with your email credentials.

2. Start the Flask Server

   python3 app.py

3. Access the Web Application

   Open your browser and navigate to http://localhost:5000.

Contributions are welcome! If you'd like to enhance InsightSummarizer, please submit a pull request or open an issue. Your feedback and support are greatly appreciated.

InsightSummarizer is released under the MIT License.

Made by Xiaoqi Weng from Birmingham with Love❤️