project.révoir : Searching & revisiting previously viewed web pages is a common yet inconvenient task for most users due to the voluminous amount of personally accessed information on the web. The complexity and the dynamic nature of the web has long outgrown the diminutive browser history feature. Searching and sorting through modern web history is an incredibly inefficient and time-consuming process. project.révoir uses supervised machine learning to leverage the natural recall process imbibed in humans of using episodic and semantic memory cues to facilitate recall, and presents a personal web history revisitation technique called révoir. This surfaces web pages through context and content keywords in the form of flash cards, tags and flags. The process takes place in three main stages Sourcing, Extraction & Analysis. The dataset which consists of the user’s web history is sourced. This dataset which consists of URLs is put through a process of ML-based content extraction which outputs a structured body of text or content attributes. The body of text is then passed on to a ML-based Natural Language Text Classifier to be tagged by content and context rather than chronology. In conclusion, our project is a ML-powered feed which organizes web history with enough context to aid the making of associations while retaining enough structure for users to navigate their collections seamlessly.

Role:
Coding, UI/UX