Читать книгу Data Mining and Machine Learning Applications - Группа авторов - Страница 20

It is a warehouse which means it collects data from multiple heterogeneous sources. It supports analytical data processing and helps in decision-making. As data is collected from various sources, before storing this data into the warehouse (Table 1.1), data cleaning, data integration, and data consolidation, etc., steps must be performed and represented in Figure 1.3 [18]. Data warehouse properties are as follows:

Table 1.1 Comparison in a data warehouse—OLTP.

Figure 1.3 Data warehouse.

 Subject-oriented—designed for a specific subject/s

 Integrated—integrates different data from multiple sources.

 Non-volatile—data once stored remains stable and does not change over time.

 Time-variant—it looks at change over time.

One can compare data warehouse and OLTP as follows: