Decomposing Human Behavior Into Quantified Layers of Perception

摘要

A method, computer program product, and system are provided for decomposing human behavior into quantifiable layers of perception. The method for decomposing human behavior into quantifiable layers of perception may be referred to as “DAP-E,” which is an acronym representative of four main phases of the method. Theses phases include the following: decompose; aggregate; propagate; and, evaluate. For example, an embodiment of the method may include aggregating one or more outputs from one or more behavioral models to represent a statistically significant conclusion and propagating the conclusion across a social network to allow opinions of a first entity within the social network to influence a conclusion of a second entity. The method may also include updating the one or more behavioral models based on new data. In aggregating the one or more outputs, the one or more behavioral models, which may be from a field of psychology, may be correlated to one or more layers of perception relating to a topic of interest. Further, the outputs may be aggregated using a Bayesian belief network to calculate an impact that the one or more outputs have on beliefs of a decision maker. A token passing behavior algorithm may be used to weigh an opinion of an entity receiving a token when propagating the conclusion across the social network. An increase in a percent confidence score from the Bayesian belief network may result from an aggregation of updated behavioral model outputs.