Specification definition involves utilizing business object modeling to construct a bus matrix and to categorize and define subject areas (data domains), business activities, business objects, atomic metrics, business filters, periods, and derived metrics. This topic outlines the process of specification definition within this tutorial.
Terms
Business Entity:
Business Object
A business object serves as the foundation for measurement, representing a specific business property. These properties collectively constitute a business object, also known as an entity object. Defining business objects is crucial when segmenting subject areas and constructing a bus matrix, and should be done in correlation with the analysis of business activities. In this tutorial, products and customers are examples of business objects.
Business Activity
Business activity encompasses the various events that occur within an enterprise, such as ordering, payment, and refunds. These activities represent indivisible behavioral events within the broader scope of enterprise operations. In this tutorial, the business activity of placing an order is exemplified by order_buy.
Atomic Metric
An atomic metric is an indivisible business measure associated with a specific business event or action, possessing a clear business significance. It is represented as
Atomic metric = Business activity (action) + Measure, such as the amount of a payment (event) transaction (measure). In this tutorial, payment_amount is an example of a measure that, when aggregated, forms the atomic metric Sum(payment_amount).Business Filter
A business filter defines the scope of business for statistical purposes. To ensure uniformity, standardization, and normative construction of all statistical metrics, a business filter is uniquely associated with a segment of the business and tied to a source logical table. The calculation logic is based on the fields within the source logical table model. In this tutorial,
product_category='West Lake Longjing'serves as a business filter.Derived Metric
A derived metric extends the business statistical scope of an atomic metric and is defined as
Derived metric = Atomic metric + Business filter + Statistical period + Business object (statistic granularity). In this tutorial, the derived metric is represented by Last 1 day_Total order amount_West Lake Longjing.