PolyBot — FAQ Chatbot

PolyBot is an offline, NLP-powered college/university FAQ chatbot built using Flask, NLTK, and Scikit-Learn. It parses and maps user questions to a pre-defined database of over 40+ FAQ entries using TF-IDF vectorization and cosine similarity matching, requiring zero external paid APIs.

How It Works

1. User Query Input: The user asks a question in the chat interface.
2. Text Preprocessing: The backend uses NLTK to tokenize, lowercase, strip stopwords/punctuation, and lemmatize the input string (e.g. converting "admissions" and "admitted" to their root word "admit").
3. TF-IDF & Vectorization: Both the user question and the FAQ database questions are vectorized into a bag-of-words model using Scikit-Learn's TfidfVectorizer.
4. Cosine Similarity Matching: Cosine similarity is computed between the query vector and all FAQ question vectors.
5. Threshold Filter: If the highest similarity score is above 0.2, the chatbot returns the corresponding answer. Otherwise, it falls back to a friendly message prompting the user to rephrase.

---

Tech Stack

• Backend: Python (Flask)
• Natural Language Processing (NLP):
  • NLTK: Tokenization, Stopwords filtering, and WordNet Lemmatization
  • Scikit-Learn: TF-IDF Vectorization and Cosine Similarity
  • NumPy: Matrix index math
• Frontend: Semantic HTML5, CSS3 (Glassmorphism & Flexbox), and Vanilla JS

---

Folder Structure

CodeAlpha_FAQChatbot/
├── app.py
├── faqs.py
├── chatbot.py
├── templates/
│   └── index.html
├── static/
│   ├── style.css
│   └── script.js
├── requirements.txt
└── README.md

---

Features

• High-Quality Custom Dataset: Built-in repository of 41+ Q&A pairs covering admissions, fees, hostel rooms, exams, library access, placements, and campus activities.
• WhatsApp/ChatGPT-Style UI: Modern dark navy theme featuring a glowing send button, animated typing indicator, and responsive bubble layouts.
• Auto NLTK Downloader: Automatically downloads NLTK packages (punkt, stopwords, wordnet, and omw-1.4) on the first run.
• Enter-to-Send: Supports standard keyboard triggers for quick sending.
• Scroll Memory: Auto-scrolls to the newest message bubbles.
• No API Costs: Executes entirely offline with no usage caps.

---

Installation & Running

Step 1: Navigate to the Project Folder

Open your terminal and enter the project folder:

cd CodeAlpha_FAQChatbot

Step 2: Install Required Python Packages

Install the required packages listed in requirements.txt:

pip install -r requirements.txt

Step 3: Run the Flask Web Server

Start the local server:

python app.py

On first run, the terminal will log the NLTK downloads. Once completed, the server will launch.

Step 4: Open in Web Browser

Open your browser and visit:

http://127.0.0.1:5000/

---

Author

Jenicadeva Christa S (CodeAlpha AI Intern)