Abstract:

The big data resources possessed by telecom operators are usually distributed in many different systems,such as DPI、OIDD、CRM.Moreover,the formulation,interpretation and rules of the big data are not always the same in different systems.Therefore,it is difficult to identify and utilize the same object’s multi-type data in different sys-tems.Big data analysis’ sample size and dimension are limited,with the decreasing of analysis results’ reality and accuracy.The methods,architectures and implementation examples of big data’s heterogeneous association were pre-sented.The data fusion in user-dimension from different systems could optimize the data sample space of applications,such as user portrait.Thus,the value of carrier’s big data density was greatly improved.

Key words: big data, telecom service big data, multi-source and heterogeneous, heterogeneous association