About Project

Users often need a centralized virtual assistant that can handle small tasks like reminders, greetings, simple queries (date/time/weather), or custom commands. Existing solutions (like Alexa or Siri) are either voice-only or not customizable for personal needs, and most bots cannot be extended easily by developers.

Problem Statement

Users often need a centralized virtual assistant that can handle small tasks like reminders, greetings, simple queries (date/time/weather), or custom commands. Existing solutions (like Alexa or Siri) are either voice-only or not customizable for personal needs, and most bots cannot be extended easily by developers.

Objective

• Build a customizable personal assistant chatbot
• Capable of handling basic day-to-day queries and actions

• Responds to date/time queries
• Manages to-do reminders
• Handles casual greetings and small talk

• Intent-based architecture for easy extensibility
• New features and tasks can be added with minimal changes

• Grammar correction for user inputs
• Includes NLP preprocessing (e.g., lemmatization, stopword filtering)

• Runs on a local server
• Can be integrated into a Django interface

Proposed Solution

• Used rule-based matching with keyword sets and regular expressions

• Integrated TextBlob for grammar and spelling correction

• Designed a custom parser to interpret commands like:
• “Remind me to call mom at 6pm”
• “Add buy milk to my to-do list”

• Responds based on user profile, previous greetings, and ongoing conversation context

• Supports optional storage and retrieval of past user queries
• Useful for session-based personalization

Technologies Used

• TextBlob – for grammar correction and text cleaning
• datetime – for handling date and time queries
• os, platform – for fetching system-level information
• re (regex) – for command parsing and parameter extraction

• Django – integrated as the primary web app interface

Challenges Faced

• Parsing human language like “remind me at 5pm” was ambiguous
• Used regex combined with rule-based templates for robust extraction

• Needed modular functions for easy, plugin-like task addition
• Designed system with intent-handler mapping for scalability

• Users typed shorthand or incorrect English
• Solved using grammar correction with TextBlob and fuzzy keyword matching

• Handled semantic understanding for phrases like “tomorrow evening” vs “at 7pm”
• Used datetime parsing with fallback defaults for incomplete input

Methodology

• Converted text to lowercase
• Punctuation removal for clean input
• Applied TextBlob for grammar correction

• Used rule-based matching with priority order
• Example priority: “weather” > “time” > “greeting”

• Executed mapped functions like:
• get_time()
• set_reminder()
• tell_joke()

• Generated human-like replies using response templates
• Used dynamic content injection for personalization

• Interactive chatbot built with Streamlit
• Django or CLI used as fallback for testing

Result / Outcome

• Personalized Greetings – Greets the user by name (configurable)
• Time & Date Reporting – Tells the current time and date on request
• Simple Reminder System – Sets reminders that are either locally stored or printed
• Custom Small Talk – Responds to casual inputs like “Who made you?” or “How are you?”
• Self-Learning Ability – Can learn new responses via an admin interface or hardcoded functions
• Local Memory Support – Optionally integrates with SQLite for storing logs or session data