DB, ETL, DW, OLAP, DM, BI relationship structure diagram

Reprinted from: http://www.cnblogs.com/ycdx2001/p/4538750.html

--------------

In the leader said the urine is not wet and the beer story, here see the original text.

(1) db/database/Database --This refers to the OLTP database, the online things database, to support production, such as the supermarket trading system. DB retains the latest state of data information, only one state! For example, every morning to get up and face in the mirror, see is the state, as for the previous day of the state, will not appear in front of your eyes, this is a db.

(2)dw/data warehouse/Data Warehouse --here is the state of the different time points in the DB, for example, every morning after washing the mirror, take a picture, every day, these photos into an album, then you can see the status of each day, This album is the Data Warehouse, he saved the data at different time points of the state, the same data information, the state of keeping different points of time, it is convenient for us to do statistical analysis.

(3)etl/extraction-transformation-loading--for the completion of the DB to DW data dump, it will be in the db of a point in time state, "extract", according to the DW storage model requirements, "conversion" Data format, and then "load" to the DW process, it should be emphasized that the DB model is ER model, conform to the paradigm design principle, and DW data model is a snowflake structure or star structure, with the topic-oriented, problem-oriented design ideas, so the DB and DW model structure is different, Conversion is required.

(4)olap--Online Analysis system , simply said is the report system, sales reports, statistical reports, and so on, this everyone is familiar with, of course, the statistics of OLAP should be more complex and richer, such as cutting, drilling and so on.

(5)dm/data mining/Data Mining -this excavation, not a simple statistic, he is based on probability theory or other statistical principles, the DW in the large data analysis, to find out we can not intuitively discover the law, for example, if we take pictures every morning, Volume of the time, also recorded the head of the day to eat things, cucumber, pig legs, roast duck, and mood, if recorded 10 years, the formation of 3,650 days of appearance and eating mood data, we each recorded, 200,000 people recorded, then, we may through these records, can be analyzed, Physical appearance and diet of the objective law; Another typical example is the British supermarket, after accumulating a lot of data, after the data analysis mining, got a rule: put the baby's diapers and beer together, sales will be better-business experts after the conclusion, careful analysis, know the reason, Because the British men like to watch more football, the wife of the child introduced a man in charge, the child urine needs diapers, and men watching football like drinking, so two commodities have a close relationship, put together sales will be better!

(6) bi/business intelligence/Business Intelligence -leaders, decision-makers, after obtaining the statistics of OLAP, and the scientific laws obtained by DM, make appropriate adjustments to production, for example, ordering supermarket staff to put beer and diapers to sell together, which is counterproductive to the DB modified inventory data-This is the role of the whole bi!

Information technology after nearly 20 years of vigorous development, many industries have accumulated a lot of valuable data, the real Big data era has come, but also gradually reflects the importance of db-etl-dw-olap-dm-bi this line, I hope you can understand the value of this big data era, to seize the opportunity, there is a better future!

DB, ETL, DW, OLAP, DM, BI relationship structure diagram