• Version
• Download 9
• File Size 510.79 KB
• File Count 1
• Create Date 12 June 2025
• Last Updated 12 June 2025

INBOX BOT: DYNAMIC SCREEN-BASED DATA EXTRACTION AND CLOUD-FREE ARCHIVING TOOL

Authors:

^1st P. Rajapandian, 2^nd A Abishaya

Associate Professor, Department of computer Applications, Sri Manakula Vinayagar Engineering College

(Autonomous), Puducherry 605008, India

Post Graduate student, Department of computer Applications, Sri Manakula Vinayagar Engineering

College (Autonomous), Puducherry 605008, India abishaya12a@gmail.com

*Corresponding author’s email address: abishaya12a@gmail.com

ABSTRACT: This project presents a semi-intelligent system that automates the retrieval of text- based content from a web-based interface (such as Gmail or similar dashboards), processes the visual data using Optical Character Recognition (OCR), and stores it into a structured database for archiving, analytics, and reporting. The solution simulates human interaction using Python automation tools and bypasses the need for API-level access, making it adaptable to a variety of use cases such as document inboxes, web mail clients, and internal portals.

The automation pipeline involves opening the web interface in a browser, capturing the screen, and identifying key visual triggers (e.g., organizational codes, transaction keywords, or date ranges). Using PyAutoGUI, Tesseract OCR, and OpenCV, the system analyzes the layout and extracts relevant content. The extracted text is processed, tagged with metadata (such as timestamp and content type), and stored as .txt files. These files are later read by a backend PHP module, parsed based on delimiters, and inserted into a MySQL database hosted on a local XAMPP server.

This modular approach makes the system highly reusable across different domains where direct content access is restricted. It enhances traceability, enables real-time monitoring, and supports secure storage of time-sensitive information without altering the source environment. By using screen-based OCR and visual recognition, the system bypasses the need for backend integration, making it ideal for restricted or legacy platforms that do not expose direct programmatic access. This also reduces dependency on external services and ensures full control over data flow, privacy, and security. Furthermore, the system is designed to operate with minimal human intervention, which reduces the chances of manual errors and increases overall operational efficiency. The use of timestamping and categorization in the structured output also enables historical tracking and facilitates audit trails—particularly useful for applications in banking, compliance, or customer support environments. Overall, this semi-intelligent, visually driven automation tool blends simplicity, adaptability, and functional precision—serving as a reliable alternative to more complex or less secure data integration pipelines.

Keywords: Screen Automation, OCR (Optical Character Recognition), Email Content Extraction, Python Automation, Web Interface Parsing, PyAutoGUI, Tesseract OCR, OpenCV, PHP-MySQL Integration, Data Archiving, Non-API Email Processing, Gmail Automation, Visual Data Retrieval, Structured Data Storage, Local Server (XAMPP), Automated Text Extraction, Clipboard Parsing, Backend Scripting, Keyword-Based Detection, Screen-Based Data Mining