Structured and unstructured data

Data comes in different formats: an image is a collection of pixels, a video is made of frames, a fee slip has a few numeric and text entries, and chat messages mix text, emoticons, images and video. By format, data is structured or unstructured.

Structured data is organised in a well-defined format, usually a table of rows and columns:

• Each column holds one attribute (also called a characteristic or variable).
• Each row holds one observation: the values of all attributes for one item.
• In the book's Table 5.1 (kitchen items in a shop), the columns are ModelNo, ProductName, Unit Price, Discount(%) and Items_in_Inventory. With a spreadsheet the owner can add the Items_in_Inventory column (128 items), or multiply each Unit Price by its Items_in_Inventory and add the results to get the value of the stock.
• More examples: books at a shop (BookTitle, Author, Price, YearofPublication); school fee deposits; ATM withdrawals (AccHolderName, AccountNo, TypeofAcc, DateofWithdrawal, AmountWithdrawn, ATMid, TimeOfWithdrawal).

Unstructured data has no fixed row-and-column structure:

• A newspaper page changes layout every day; an email can have any number of lines and attachments.
• Other examples: web pages with text and multimedia, text documents, business reports, books, audio and video files, social media messages.
• This book handles structured data only.

Metadata is data about data, often used to describe unstructured data. An email's subject, recipient, body and attachments are its metadata; an image's size (KB or MB), type (JPEG, PNG) and resolution are the image's metadata.

Example (new): a library's book register with columns for accession number, title, author and date issued is structured data; the scanned photos of old handwritten letters in its archive are unstructured, though each photo's file size and date taken are metadata.