In the realm of Business Analysis, data modelling is a critical skill that intersects our work frequently. This necessitates a foundational understanding of data modelling for every business analyst. In this piece, I aim to demystify the core concepts of data modelling and offer an example to deepen your comprehension.
At the heart of “data modelling” lies two elemental concepts: 'data' and 'modelling'. Data is omnipresent, encompassing anything you might document as information, revisit for subsequent use, or mention in dialogue. It is the lifeblood of all systems on our planet. Whether it's the capture of an employee's personal details in a corporate database or the particulars of a transaction, it represents data in action.
The term "information" signifies "data endowed with meaning". A rocket scientist may view numerical data sets as meaningful information, whereas others may simply see them as data without inherent meaning, except for their numerical nature.
Data modelling is concerned with identifying the data we aim to collect for a particular purpose and organizing it logically. It's not enough to gather data; we must weave it into a tapestry that, when viewed as a whole, offers meaningful insights.
Imagine a scenario where a company pays an employee monthly based on their role. The challenge is to encapsulate all pertinent information about the employee's role, salary, and payment details. This necessitates a database designed to house this wealth of data.
The journey of data modelling begins with defining the various 'entities' or 'concepts' — the items about which data will be collected. Entities can be anything tangible, like people, places, or products.
Consider the following entities for our illustrative example:
Employee (the individual)
Job Role (the position within the organizational hierarchy)
Salary Package (the terms of compensation)
Payment Details (the banking information)
Identifying the entities is just the start; next, we determine the attributes, the specific details, we want to record for each entity. For instance, for an 'Employee', we may collect details such as Name, Addresses, Date of Birth, City, Country, and Phone Number.
Advanced Concept: It's crucial to collect only the data that is pertinent to each entity to avoid redundancy. This efficiency is a guiding principle in data modelling.
After pinpointing the entities and attributes, the next phase is to outline the interrelations among these entities. These relationships are pivotal and should reflect how entities interact with one another. In our example, an Employee 'performs' a Job Role, illustrating a direct relationship.
Determining the 'cardinality' or 'multiplicity' of these relationships is essential. This involves assessing the permissible or necessary number of linkages between entities. For example, an Employee is associated with one Job Role, but a Job Role may be associated with multiple Employees.
The final stride in data modelling is delineating the meta data. Meta data is essentially data about data, offering insights into the use and origin of your data. For example, it could provide details about when a record was added, who added it, and its accuracy.
With this overview of data modelling fundamentals, as a Business Analyst, you're now equipped to navigate and interpret data models with confidence.