Memory-Forgetting Model based on Virtual and Real Spaces for Commercial Recommendation

Abstract

With the social media development, semantic web and Crowdsensing blog platforms become the store of big data, which store the memory of diverse information. Memory enables past experiences to be remembered and acquired as useful knowledge to support decision making in commercial recommendations, especially when perception and computational resources are limited in Crowdsensing. However, the traditional event research lacks the view of memory and forgetting of event from the people's cognitive psychology. On the basis of Ebbinghaus, we treat every day's information as an independent event to calculate Memory-Forgetting Curve (MFC), then calculate MFC of episode with the fusion of events'.