Storm-based optimization method of the food data parallel-calculation consistency problem

The invention discloses a Storm-based optimization method of the food data parallel-calculation consistency problem. First, food event data are acquired, then single-pass clustering is carried out oneach event, then the problem of data inconsistency is solved through two manners of dynamically cluster increment acquisition and delay calculation, and thus stream-type data clustering is realized. The method is used to discover public opinion orientation, and quickly respond to a public concerning point. According to the Storm-based optimization method of the food data parallel-calculation consistency problem provided by the invention, processing efficiency and accuracy of public opinion data are improved through parallel processing of a Storm distributed-framework and methods of dynamic cluster increment acquisition and random delay, and the problems of parallel-processing data inconsistency and excessive cluster traversal frequency are solved. Therefore, the technical solution providedby the invention realizes