Research on Big Data Processing Based on Relational Logic

Improving the speed of system operation is an effective way to solve the problem of Big data, and the most direct method is to perform high performance computing from the original value layer. The method of optimization design is proposed based on the speed, function, resources and the connection of the programmable components, the operations in the same format are merged into a class of relation, the relation table on the input corresponding to the output data is listed as a function of the constraint, each column shows the a bit field and the data in the rows is determined by the function. According to the relationship, the logical expression of the output value is obtained using a domain, neighbourhood, global operations, differential time control can achieve for the number or bit length of different data changes. The design methods of two logic relations with accumulation and decision are given and it is applied to address transform and mining analysis on Big data.