纯英语数据仓库指南[+示例]

为了做出明智的决定，公司需要准确且可访问的数据。“可访问”部分是关键：如果数据不集中，整个组织都可以使用，您将获得data silos。筒仓非常适合耕种……不过，对其他企业而言并不是那么多。

A big reason for data silos is a lack of automation — data needs to be moved manually from one source to a database, application to application, et cetera. This friction promotes a culture of siloed data, which limits a business’s ability to make data-driven decisions, which inhibits growth. Plus, given the sheer amount of raw data an enterprise company can generate, manual data sharing quickly becomes not just impractical, but impossible.

为了解决这个问题，我们可以转向数据仓库。数据仓库创建了一个可扩展且功能强大的系统，在该系统中，数据自动处理并与适当的当事方共享。通过将数据组织到一个位置，您的员工可以更快地解决问题，并始终如一地完成期限。

由于数据仓库通常与其他类似概念相混淆数据管理，让我们定义它是什么，并确定一些不同类型的数据仓库。

数据仓库概念：什么是数据仓库？

A data warehouse is a business intelligence system that brings together large volumes of data from multiple sources into a centralized repository for more efficient organization, analysis, and reporting. Sources could include website data capturing tools, purchases and transactions, inventory tracking systems, an企业资源计划系统（ERP）, and marketing and sales software.

数据仓库的主要工作是将所有业务系统生产的大量数据合成一个可访问的位置。在数据仓库中，信息在分析师进行审查的同时连续流动。这使得企业可以创建报告和仪表板，以不断监视和改善业务功能。

Image Source

Data warehouses are not a new concept, but have grown more sophisticated with the rise of cloud technologies. From low-level to high-level, a data warehouse usually includes a database to hold the raw data, software to extract data from the database and prepare it, and tools for analysis,报告和数据可视化。数据仓库还可以应用高级AI和机器学习技术。

What are the benefits of a data warehouse?

The main purpose of a data warehouse is to aggregate a business’s data assets into a single source of truth for analysis and insights, instead of requiring analysts to track down data from dozens of siloed sources. Additional benefits include:

• More actionable information:数据仓库不仅汇编了大量数据，而且还促进了分析和报告，以帮助最终用户识别大型图片趋势并做出扎根于数据的决策。
• 改进数据的完整性:将所有数据放在一个地方都是一个良好的开始，但是如果您处理不准确的数据，这是不值得的。数据仓库可以标准化和清洁其数据，因此您可以确保处理可靠的信息。
• 实时数据：数据仓库实时更新您的业务最准确的情况。
• 改进performance:数据仓库旨在处理非常大的数据集，而没有滞后时间或对公司其余技术基础架构的影响。raybet电子竞技即使数据量扩展，数据仓库也可以非常快速地操纵数据。
• AI：Data warehouses are capable of integrating artificial intelligence and machine learning with their datasets to uncover hidden patterns via数据挖掘。人工智能是主要的竞争优势，因为它在商业智能中变得更广泛地采用。

有关数据仓库概念的更多说明，请查看365个数据科学的视频，该视频涵盖其其他定义功能：

Data Warehouse vs. Database

It’s important to note that data warehouses are different from databases. While both store data, their purposes differ significantly.

数据库是将数据组织到行和列中的结构，从而使信息易于阅读。与数据仓库相比，数据库是仅用于存储的简单结构。

数据仓库可能由许多数据库组成。数据仓库通过从多个来源收集数据并允许数据分析来超越简单的数据库。数据仓库不仅存储数据 - 它们将其汇总用于长期业务使用。

数据仓库与数据湖

You may have also heard of “data lakes.” A data lake also stores raw data from different sources, but this data hasn’t been filtered or structured. When businesses keep a data lake, they usually intend to use this data later in a more structured manner.

另一方面，数据仓库存储已过滤，清洁和定义的供特定用途的结构化数据。数据仓库是用于分析和提取见解的，因为它们所包含的数据更为可行。

Because data lakes include raw data, the data is simpler to use and easier to edit. In data warehouses, the data is more expensive to make changes to, but better for long-term decisions.

Image Source

数据仓库的类型

企业数据仓库

企业数据仓库是中央数据库，在该数据库中，将数据组织，分类和用于决策。这些系统还将标记数据并将其分类，以便于访问。

Operational Data Store

尽管企业数据仓库更适合长期业务决策，但对于日常的日常活动而言，使用运营数据存储（ODS）是优选的。ODS将在实时更新，并存储特定于所选活动的数据。

数据库

数据集市是支持特定业务部门，团队或职能的数据仓库的一部分。通过数据mart的任何信息都会自动存储和组织以供以后使用。数据集市具有与数据仓库相同的好处和功能，只是较小的规模。

现在你已经熟悉的基本面data warehouses, let's take a look at some common concepts used by most businesses.

3 Data Warehouse Concepts with Examples

Basic Data Warehouse

A basic data warehouse aims to minimize the total amount of data that is stored within the system. It does this by removing any redundancy within the information, making it clear and easy to look through.

As you can see in the example below, this concept centralizes information from a variety of sources. Employees then access data directly from the warehouse. This system is useful for SMB who want a simple approach to data storage.

Image Source

Data Warehouse With Staging Area

Some data warehouses clean and process data before moving it into storage. These systems have "staging areas" where information is reviewed, evaluated, then deleted or transferred into the warehouse. This ensures that only relevant and useful data are stored within the software.

如果您查看下面的示例，则可以看到登台区域位于数据源和仓库之间。对于处理大量客户数据的企业，此过程将滤除对团队无益的无关信息。

Image Source

带有数据集市的数据仓库

Data marts add another level of customization to your data warehouse. Once data is processed and evaluated, data marts streamline information to teams and employees who need it most. That makes your departments significantly more productive because customer data is being delivered directly to them.

在下面的示例中，我们可以看到如何使用数据集市向销售和库存团队发送信息。这有助于业务领导者做出更快的决策并利用及时的营销机会。

Image Source

数据仓库：保持数据的最新

数据仓库是商业智能系统的主食，为扩展公司带来了许多好处。如果您的企业面临管理大量日期并在整个团队中分配的挑战 - 同时也努力利用这些数据来获得有意义的见解，那么数据仓库可能是您的最佳选择。

Editor's note: This post was originally published in August 2019 and has been updated for comprehensiveness.