This invention describes a way for online stores to give you better product recommendations. It works by looking at your past browsing activity and ignoring any of it that happened before you bought something in that same product category. For example, if you browsed several TVs and then bought one, the system will stop using that old TV browsing to recommend more TVs, assuming you've fulfilled that need.
Why it matters: Filed before the widespread adoption of sophisticated machine learning for recommendation engines. The specific rule-based filtering described, while innovative in 2011, is now a relatively simple logical component to integrate into today's advanced AI-driven recommendation platforms.
AI gives you a few directions you could take this. Pick one, and we check whether your version is different enough to patent, then write the filing.
Reinvent this with AI