Big Data and BI data analysis What is the difference, application scenarios large data sources, technology architecture.
BI (Business Intelligence)
Chinese translation for business intelligence, is a complete solution for the existing data will be organized for effective integration, providing fast and accurate reports and make decisions based on helping organizations make informed business decisions.
Big Data
From massive amounts of data collected, these data from different sources, formats for direct analysis algorithms and find the correlation between the data. Simply put, more emphasis on big data discovery, as well as speculation and confirms the cycle of the approximation process.
1. Structured VS unstructured
2. The overall individual VS
Big Data for traditional BI, both inherited, but also development, from the "Tao" perspective, BI and big data difference is that the former is more inclined to decision-making, based on the fact that the description is more common groups to help policymakers grasp the trend of macro statistics , support for business operating metrics class of problems, big data is a broader connotation, tend to portray the individual, more personalized that decision.
3. Traditional database VSHadoop.
Traditional BI using ETL, data warehousing, OLAP, visualization reporting technology, ERP, belongs to the application and presentation layer technology, there are at the edge of elimination, because it will not solve the problem of massive amounts of data (both structured and unstructured) of . And large data application is a complete system, including problem solving ETL structured mass, with the Hadoop unstructured data, streaming technology, calculation of mass data calculation Hadoop, MPP technology, with Redis, HBASE other ways to solve the problem of efficient reading, for online analysis and other technical issues such as Impala. So a new industry.
4. The distributed storage centralized storage VS
Large data environment, the data stored on the distributed file system (e.g. the HDFS), rather than stored on a central server. The data is distributed on the working node, for ease of handling. Distributed file system more secure and flexible.
5. Reporting analytics VS exploratory discovery and analysis:
Personal understanding: a model that has been outside the box, fill it up something, and then the result. One is like Newton, Apple hit your head, you want to explore this phenomenon
Big Data scenarios:
retail
E-commerce
express delivery
