The big data between your ears: Human inspired heuristics for forgetting in databases

Inspired by explanations of human forgetting, the primary research question that motivates this research is whether databases can be manipulated to resemble human long term memory in performance. We hypothesize that selective degradation of a database using different heuristics proposed as explanations of human forgetting, produce different recall results. Furthermore, we explore how these results resemble human performance by mapping performance to five memory phenomena that typify human forgetting. For this research we implemented the artificial long term memory of a fictitious character, Mr. Alfred Polly, using the Ardemia modelling architecture, which is built on a relational database management system. The experimental investigation of three versions of Mr. Polly´s selectively degraded artificial memory revealed that forgetting heuristics produce query results that resemble human forgetting. The implication of this study is that big data algorithms inspired by “human forgetting inspired heuristics” can be a tool for shrinking and managing large data using attributes or metadata.