Bi stands for business inteligence. This helps enterprises use data more effectively and provide business decision-making support. It allows decision-making managers to obtain key information anytime and anywhere. Based on Digital decision-making, the decision-making level is improved.
It includes data reports (traditional data reports and core data reports), OLAP (online link analysis), and data mining. It is a comprehensive application of data warehousing, OLAP, and data mining technologies.
Bi solution:
DB-> ETL-> data warehouse-> Data Mining | OLAP-> dataanalyzing | datareporting
-> Bi Portal
ETL: Extract, transfer, local
A data warehouse is an independent data environment. You need to import data from the OLTP (online transaction processing) Environment, external data sources, and offline data storage media to the Data Warehouse through the extraction process (ETL.
ETL is used to extract data from distributed heterogeneous data sources (such as relational data and flat data files) to the temporary middle layer for cleaning, conversion, and integration, finally, it is loaded into a data warehouse or even data, which becomes the basis of OLAP and data mining.
Extraction: heterogeneous data sources; Conversion: filtering data, Processing Inconsistent data according to established rules; loading: writing data to a data warehouse or data mart.
SSIS (Microsoft SQL server integration services), a platform that claims a high-performance solution (including ETL for data warehouses), a visual programming environment without coding.
CDC: Change Data Capture
Monitors data updates, captures updates to business data tables, and implements incremental updates to data warehouses.
Solution: 1. Add a special identifier column to the data table.
2. Create a trigger on the data table.
3. SQL Server Change Data Capture: in a relational database, all data operations are recorded in the log, and the complete data operation history can be obtained through log analysis.
OLAP & cube (CORE)
OLAP :( Online Analytical Processing): enables analysts, administrators, or executors to convert from multiple perspectives to raw data, which can be truly understood by users, it also truly reflects the enterprise's multi-dimensional features for fast, consistent, and interactive storage, so as to obtain a type of software technology with a deeper understanding of data.
Cube can be understood as a multi-dimensional database.
Dimension: the specific angle of observation data, such as time and geography.
Level: a detailed description of a specific angle.
Dimension member (member): the value of a dimension.
Measure: The observed object and value. Such as sales and sales.
Business Intelligence (BI)