METHOD FOR CLOUD BASED REAL-TIME PROCESSING BIGDATA STREAM

摘要

Provided is a real-time processing method of a cloud based big-data stream, comprising: a step of assigning a big-data stream to a neuron having at least one node; a step of checking defects based on a level of detected fraud data if the fraud data is detected among data included in the big-data stream in the assigned neuron; and a step of assigning the fraud data to a danger zone if the level of the fraud data corresponds to a predetermined level, and outputting an alert with respect to the fraud data which is re-determined by being assigned in the danger zone.