Purpose

Automatically summarize long documents, news articles, research papers, emails, or reports into short, meaningful summaries using NLP and transformer models.

Used In Real Life By

• Content creators
• Students & researchers
• News agencies
• HR (summarizing resumes & job descriptions)
• Corporate teams (summarizing long reports)

---

🧠 What This Project Will Do

✔ Accept text from:

• PDF
• DOCX
• URL
• Plain text

✔ Clean and preprocess the content
✔ Summarize using AI models
✔ Output multiple styles of summaries:

• Short summary
• Detailed summary
• Bullet-point summary
• Title generation

✔ Save results to a text file or JSON
✔ Optional GUI or REST API

---

🧰 Tech Stack

• transformers (HuggingFace models)
• PyPDF2 (PDF)
• python-docx (DOCX)
• BeautifulSoup + requests (web pages)
• pandas / json (output)

---

📁 Folder Structure

AI_Text_Summarizer/
│── summarizer.py
│── input/
│     ├── sample.pdf
│     ├── sample.docx
│── output/
│
└── models/   (optional)

---

🔥 HuggingFace Summarization Model

We will use:

facebook/bart-large-cnn

Best quality & fast.

---

🧠 Full Python Code: summarizer.py

import PyPDF2
import docx
import requests
from bs4 import BeautifulSoup
from transformers import pipeline

# Load AI summarization model
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

# --------- FILE READERS ---------

def read_pdf(path):
    text = ""
    with open(path, "rb") as f:
        reader = PyPDF2.PdfReader(f)
        for page in reader.pages:
            page_text = page.extract_text()
            if page_text:
                text += page_text + " "
    return text

def read_docx(path):
    doc = docx.Document(path)
    return " ".join([p.text for p in doc.paragraphs])

def read_url(url):
    html = requests.get(url).text
    soup = BeautifulSoup(html, "html.parser")
    return soup.get_text(separator=" ")

# --------- AI SUMMARIZER ---------

def make_summary(text):
    # HuggingFace models accept max ~1024 tokens, so summarize in chunks
    chunk_size = 1000
    chunks = [text[i:i+chunk_size] for i in range(0, len(text), chunk_size)]

    outputs = []
    for chunk in chunks:
        summary = summarizer(chunk, max_length=150, min_length=60, do_sample=False)
        outputs.append(summary[0]["summary_text"])

    final_summary = " ".join(outputs)
    return final_summary

# --------- MAIN ---------

def summarize_any(path_or_url):
    if path_or_url.startswith("http"):
        text = read_url(path_or_url)
    elif path_or_url.endswith(".pdf"):
        text = read_pdf(path_or_url)
    elif path_or_url.endswith(".docx"):
        text = read_docx(path_or_url)
    else:
        text = open(path_or_url, "r").read()

    text = text.strip().replace("\n", " ")

    summary = make_summary(text)

    # Bullet points
    bullets = "\n".join([f"• {line.strip()}" for line in summary.split(".") if line.strip()])

    # Title generation
    title = summarizer(summary, max_length=20, min_length=5, do_sample=False)[0]["summary_text"]

    return {
        "title": title,
        "summary": summary,
        "bullet_points": bullets
    }


# --------- RUN EXAMPLE ---------

if __name__ == "__main__":
    result = summarize_any("input/sample.pdf")

    with open("output/summary.txt", "w") as f:
        f.write("TITLE:\n")
        f.write(result["title"] + "\n\n")
        f.write("SUMMARY:\n")
        f.write(result["summary"] + "\n\n")
        f.write("BULLET POINTS:\n")
        f.write(result["bullet_points"])

    print("Summary saved to output/summary.txt")

---

📌 Example Outputs

Title Generated:

“Impact of AI on Modern Businesses”

Short Summary:

AI technologies significantly improve business efficiency by automating repetitive tasks, optimizing decision-making, and enhancing customer experience…

Bullet-Point Summary:

• AI increases operational efficiency  
• Automates repetitive tasks  
• Enhances customer experience  
• Enables better data-driven decisions  
• Popular in finance, healthcare & retail  

---

🚀 Advanced Enhancements

🔹 1. Add GUI with Tkinter / PyQt

Drop file → Get summary instantly.

🔹 2. Make a REST API

Use FastAPI → /summarize endpoint.

🔹 3. Chrome Extension

Right-click → “Summarize this page”.

🔹 4. Multi-language summarization

Add multilingual models (MBART).

🔹 5. PDF export of summary

Integrate with ReportLab.