User events/behaviors and perceptual computing system emulation

摘要

Methods, apparatuses and storage medium associated with engineering perceptual computing systems that includes user intent modeling are disclosed herewith. In embodiments, one or more storage medium may include instructions configured to enable a computing device to receive a usage model having a plurality of user event/behavior statistics, and to generate a plurality of traces of user events/behaviors over a period of time to form a workload. The generation may be based at least in part on the user event/behavior statistics. The workload may be for input into an emulator configured to emulate a perceptual computing system. Other embodiments may be disclosed or claimed.

主权项

1. At least one non-transitory computer-readable storage medium having a plurality of instructions to enable a computing device, in response to execution of the instructions by the computing device, to:
receive a usage model having a plurality of user event/behavior statistics; generate a plurality of traces of user events/behaviors over a period of time, based at least in part of the on the user event/behavior statistics of the usage model, form a workload from the plurality of traces for input into an emulator that emulates a perceptual computing system; and output the workload to the emulator to enable evaluation of performance of the perceptual computing system based on the workload; wherein the emulator emulates the perceptual computing system; wherein evaluation of performance includes evaluation of power consumption or latency of hardware components of the perceptual computing system; wherein for evaluation of performance, the emulator is to receive a plurality of component models of the perceptual computing system that specify power consumption or latency behaviors of corresponding hardware or software components of the perceptual computing system; wherein the emulator is to further receive a plurality of usage modality algorithms to recognize usage modalities from output data of the component models; and generate and output emulated responses of the perceptual computing system to the workload, using the component models and the usage modality algorithms; and wherein the emulated responses include corresponding power consumption or latency information of the hardware or software components of the perceptual computing system associated with processing the workload.