Abstract:
With high rate of internet penetration in Bangladesh, all the people are increasing their presence online and so is the case with phishing attacks threatening day to day for everybody, specially ones who do not belong to any sort of technical background. Many of the available tools are resource demanding and require online connection, thus rendering them too heavy to implement on legacy devices. This work has addressed this missing by developing an illustrated anti-phishing victim-assistance system:Banglar Shahzada Phishing Guard as a light-weight, bilingual Bangla and English ideal of privacy focused web application. Instead, it is written on top of a self-sufficient node. js backend with a deterministic In-Memory Hash Map Indexing algorithm. Such design decision enables the system to be able to perform a real-time URL lookup with O(1) complexity against PhishTank without being affected by any API latency. Also I've built the client side Shadow AI on my own, it hunts multi-level threats with Shannon Entropy logic and Homoglyph attacks. Performance However, preliminary testing shows that the Application responds within milliseconds to property transactions, so even users with low-end PCs Intel i3 with 4GB Ram will feel secure.
