Day 5 – Types of Data: Structured, Semi-Structured, and Unstructured | Become a Job-Ready Data Analyst in 180 Days

To become a Data Analyst, it is not enough to simply look at data. You also need to understand the format in which data is available and choose the right way to inspect it.

Sometimes data is available in rows and columns. Sometimes it comes in formats such as JSON. In other cases, it may be found in emails, images, videos, or customer reviews. These different formats cannot always be handled in the same way.

In this lesson, we will explore three important types of data:

  • Structured Data
  • Semi-Structured Data
  • Unstructured Data

By the end of this lesson, you should be able to identify common data formats and choose an appropriate starting point for inspecting each type.

1. What Is Structured Data?

Structured Data is data organized according to a predefined structure. It is generally arranged in rows and columns.

In this format, each column has a specific meaning, and each row represents a record.

Example: Online Store Orders

Order_ID Customer Product Amount
101 Rahul Keyboard 1500
102 Priya Mouse 800
103 Aman Monitor 12000

Note: This is an illustrative example, not a result from an actual business dataset.

In this table:

  • Each row represents an order record.
  • Order_ID identifies the order.
  • Customer contains the customer’s name.
  • Product identifies the purchased product.
  • Amount shows the order amount.

The information is clearly organized, so it is relatively easy to read, filter, sort, and summarize.

Common Examples of Structured Data

  • Properly organized tables in Excel or Google Sheets
  • SQL database tables
  • Employee records
  • Sales transaction records
  • Inventory tables
  • Customer information tables

How to Inspect Structured Data

When inspecting structured data, an analyst should check a few basic things:

  1. What are the column names?
  2. What does each row represent?
  3. Are the values stored in a consistent format?
  4. Are there any missing values in the columns?
  5. Are identifiers unique where they are expected to be?

Structured data can be inspected using tools such as spreadsheets, SQL databases, and Python pandas.

Remember:

The main characteristic of structured data is its predefined rows-and-columns structure. However, having a structured format does not automatically mean the data is clean or error-free.

2. What Is Semi-Structured Data?

Semi-Structured Data is data that has some organization but does not completely follow the fixed rows-and-columns structure of a traditional table.

It may use keys, values, tags, or nested objects to organize information.

In simple terms, the data is not completely unorganized, but its structure is also not the same as a normal spreadsheet table.

Example: JSON Data

Imagine that an online store’s order system stores order information in JSON format.

{
  "order_id": 101,
  "customer": "Rahul",
  "product": "Keyboard",
  "amount": 1500,
  "status": "Delivered"
}

In this example:

  • order_id is a key.
  • 101 is the value associated with that key.
  • customer identifies the customer field.
  • amount represents the order amount.

The information is clearly labeled, but it is not directly presented as a traditional spreadsheet row.

Common Examples of Semi-Structured Data

  • JSON files
  • XML documents
  • HTML documents
  • Application logs with labeled fields
  • API responses available in JSON or XML format

How to Inspect Semi-Structured Data

When inspecting semi-structured data, an analyst needs to understand its format and internal organization.

  1. First, identify the format, such as JSON or XML.
  2. Identify the keys, values, and tags.
  3. Check whether the data contains nested objects or arrays.
  4. Understand which pieces of information are relevant.
  5. If needed for analysis, convert relevant fields into a tabular format.

A text editor or suitable JSON viewer can be used to inspect JSON. Python also provides libraries for reading and processing JSON data.

Important Note:

JSON can contain nested objects and arrays. When converting it into a table, it is important to preserve the relationships between fields and their original meaning.

3. What Is Unstructured Data?

Unstructured Data is data that does not have a predefined traditional tabular structure like rows and columns.

This does not mean that the data contains no useful information. Useful information may be present, but extracting and analyzing it can require different methods.

Example: Customer Review

Imagine that a customer writes the following review on an online store:

“The product quality is good, but the delivery was later than the expected date.”

The customer has shared two different experiences:

  • The product quality was good.
  • The delivery was late.

An analyst may need to identify useful information from this text. A normal numerical calculation alone will not automatically explain the full review.

Common Examples of Unstructured Data

  • Customer feedback and free-text reviews
  • Emails and documents
  • PDF reports
  • Images and photographs
  • Audio recordings
  • Videos

How to Inspect Unstructured Data

The inspection method depends on the data format.

  • Text: Read the content and identify themes, complaints, and common topics.
  • PDF: Inspect the document’s text, tables, and relevant sections.
  • Images: Examine the image content and relevant visual information.
  • Audio: Listen to the recording or use a suitable transcription process.
  • Video: Examine relevant scenes, timestamps, and available text or transcripts.

Depending on the task, analyzing unstructured data may require techniques such as text analysis, document processing, image analysis, or audio and video processing.

Common Mistake:

It is incorrect to assume that unstructured data is useless. Customer reviews, emails, and feedback may contain important business information. The main difference is that extracting this information requires a different approach.

4. Structured vs Semi-Structured vs Unstructured Data

Let’s compare all three types side by side.

Basis Structured Semi-Structured Unstructured
Organization Predefined tabular structure Organized through labels, keys, or tags No traditional tabular structure
Format Rows and columns Key-value pairs, tags, nested structures Text, images, audio, video
Examples SQL tables, spreadsheets JSON, XML, HTML Emails, images, recordings
Inspection Columns, rows, values Keys, values, tags, nested fields Content and format dependent

5. Real Business Scenario: An Online Store

Let’s use a simple business scenario to understand how one company can work with all three types of data.

Imagine an online store that collects information related to its customers and orders.

Structured Data

The store’s order database contains fields such as order ID, customer ID, product ID, and order amount.

Possible analysis: Summarize total orders, total sales, and product-wise sales.

Semi-Structured Data

The store’s application generates JSON records related to order status and customer activity.

Possible analysis: Inspect the available fields and understand relevant order events.

Unstructured Data

Customers send product reviews, complaints, and support emails.

Possible analysis: Review customer feedback to identify common complaints and product-related themes.

Notice an important point: all three data types have different formats, but each can provide useful information about the business.

6. Why Should a Data Analyst Understand These Differences?

A Data Analyst should not start analyzing data immediately after receiving it. First, the analyst needs to understand the form in which the data is available.

Understanding the data type helps an analyst make several decisions:

  1. Inspection method: How should the data be read and understood?
  2. Tool selection: Should a spreadsheet, SQL, text editor, or another suitable tool be used?
  3. Preparation: Does the data need to be transformed or converted into a tabular format?
  4. Analysis approach: How can the required information be extracted from the data?
  5. Limitations: What limitations may arise from the available format or content?

Remember, one tool or analysis method is not suitable for every type of data.

7. Common Mistakes Beginners Should Avoid

  • Assuming structured data is automatically clean: Even data arranged in rows and columns can contain missing, duplicate, or incorrect values.
  • Treating JSON like a normal spreadsheet: Nested fields and arrays need to be understood.
  • Ignoring unstructured data: Text feedback and documents may contain useful business information.
  • Using the same tool for every format: Choose the tool and method according to the nature of the data.
  • Confusing data format with data quality: Structured, semi-structured, and unstructured describe how data is organized. Data quality is a separate concern.

8. Practice Exercise: Identify the Data Type

Task: Classify each example below as Structured, Semi-Structured, or Unstructured.

  1. An Excel sheet containing Employee_ID, Name, Department, and Salary columns.
  2. A JSON file containing customer details and an address stored in a nested format.
  3. A customer email containing a written complaint about a product.
  4. A SQL database table containing product ID, product name, and price.
  5. A video recording in which a customer is reviewing a product.
  6. An XML document containing order details organized through tags.

Expected Deliverable: Create a list with the data type and a short reason next to each example.

Validation: Check that you have identified the actual organization format, not just the file extension.

Show answers with explanations
  1. Structured: The Excel sheet contains predefined columns and rows.
  2. Semi-Structured: JSON uses keys and may contain nested address information.
  3. Unstructured: The complaint text in the email is not organized in a traditional tabular format.
  4. Structured: The SQL table follows predefined columns and records.
  5. Unstructured: Video content does not follow a traditional rows-and-columns structure.
  6. Semi-Structured: XML organizes information through tags.

The main basis for classification is how the data is organized, not how important it is to the business.

9. Quick Knowledge Check

Q1. Which type of data generally follows predefined rows and columns?

  • A. Unstructured
  • B. Structured
  • C. Semi-Structured

Q2. JSON data is generally classified as which type?

  • A. Structured
  • B. Semi-Structured
  • C. Unstructured

Q3. A customer’s free-text review is an example of which type?

  • A. Unstructured
  • B. Structured
  • C. Relational

Q4. Is structured data automatically error-free?

  • A. Yes
  • B. No
Check your answers
  1. B. Structured — it follows predefined rows and columns.
  2. B. Semi-Structured — it organizes information through keys and values.
  3. A. Unstructured — a free-text review is not arranged in a traditional table.
  4. B. No — structured data can still contain missing, duplicate, or incorrect values.

10. Key Takeaways

  • Structured Data: Organized in predefined rows and columns.
  • Semi-Structured Data: Organized through keys, values, tags, or nested structures.
  • Unstructured Data: Does not follow a traditional tabular structure. Examples include text, images, audio, and video.
  • The inspection method should be selected according to the data format.
  • Data format and data quality are two different concepts.

11. Next Lesson Preview

Day 6 – Rows, Columns, Records, Variables, and Data Types

In the next lesson, we will explore tabular data in more detail. You will learn what rows, columns, records, and variables mean, and how to identify what a single row in a dataset actually represents.

This understanding will become an important foundation for working with spreadsheets, SQL, Power BI, and Python.

Dev Sir’s Advice: One of the first important steps in becoming a Data Analyst is learning to understand data. Once you can identify the format of the data, you can choose a better approach to inspect and analyze it.

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top