AI-Powered Digital Hoarding Cleaner

Authors: Asscoiate Professor K Deepa Shree, Sanjana S, Shreya R, Sinchana Adiga, Sharanya

Abstract: In today’s data-driven world, individuals and organizations generate vast amounts of digital data, often leading to inefficient storage and digital clutter—commonly known as digital hoarding. Traditional storage cleaners primarily focus on file size or duplication and lack the intelligence to make context-aware decisions. This paper proposes an AI-powered Digital Hoarding Cleaner that leverages advanced Natural Language Processing (NLP) and Machine Learning (ML) to analyze file content and user behavior. The system integrates models such as BERT, GPT, and BART for file summarization and semantic understanding. It offers personalized recommendations to retain, delete, or reorganize files based on relevance, using techniques like semantic similarity detection, fuzzy matching, and behavioral analytics. Cloud support for services such as Google Drive and OneDrive ensures seamless integration, while a user-friendly web dashboard provides insights into storage patterns and suggested actions. Emphasizing data privacy, the tool enables local processing and secure communication. Overall, the system aims to optimize storage usage, reduce cognitive overload, and deliver intelligent file management beyond traditional cleaning methods.

DOI: http://doi.org/