Suggestions for lightweight on-device machine learning to classify promotional vs transactional notifications, requests for smarter filtering beyond regex patterns
← Back to Show HN: DoNotNotify – Log and intelligently block notifications on Android
Users are increasingly enthusiastic about on-device machine learning as a privacy-first solution for filtering intrusive promotional notifications while preserving essential transactional alerts. While some developers express concern over the computational demands of implementing ML, commenters suggest that lightweight models or local LLMs could effectively automate complex classification tasks that traditional regex patterns miss. Beyond simple blocking, there is a strong desire for advanced features like "spam inbox" bucketing and automatic coupon extraction, all while maintaining the low battery impact characteristic of current push-based processing.
5 comments tagged with this topic