Computer Science · Big Data · Topic 1 of 5

A library with ten million books isn't big data.

"Big data" doesn't just mean "a lot of data." What makes data big, in the computer science sense, is a combination of properties that break the tools and techniques used for ordinary datasets — and every 91908 Big Data question assumes you can name and apply them accurately.

AS91908 · Big Data·Vocabulary this whole topic builds on

Photo: Georgia de Lotz / Unsplash

Executive summary

Goal: define each of the "Vs" of big data precisely, and apply them to a named, real case study.

Section	What it covers
What is big data?	The three Vs (and two more), straight from the specification
Big data vs "just a lot of data"	Why the distinction changes what tools you need
Worked example	Google's 2020 NZ Community Mobility Report, all three Vs at once
Applying this to an unfamiliar context	A five-question checklist for any new scenario

What is big data?

Big data refers to datasets that are too large, too fast-moving, or too varied in structure to be efficiently stored, processed, or analysed using traditional single-computer methods (e.g. one spreadsheet, one relational database on one machine).

The standard way to describe why a dataset counts as "big" is the three Vs, named directly in the 2026 specification:

V	Meaning	Example
Volume	The sheer amount of data being generated and stored	X (Twitter) generates hundreds of millions of posts per day worldwide
Variety	Data exists in many different formats and structures at once	Numeric like-counts, timestamps, free-text captions, images, video, geolocation
Velocity	The speed at which data is generated and needs to be processed	Trending topics need near-real-time analysis to stay relevant within minutes

The specification adds "etc." after these three, and two further Vs are commonly taught alongside them:

• Veracity — how trustworthy, accurate, or noisy the data is (e.g. bot accounts, spam, deliberately false posts)
• Value — whether the data is actually useful once processed (huge volume is worthless if it can't be turned into insight)

You don't need to memorise a fixed list — the spec explicitly leaves it open — but you should be able to define each V precisely and apply it to a named scenario, not just recite it.

Big data vs "just a lot of data"

A common mistake is treating "big data" as a synonym for "large file." The distinction matters because it explains why new tools are needed at all.

A school's enrolment spreadsheet is large but not "big data" — it's one consistent structure, updated occasionally, and fits comfortably on one computer. A social media platform's post stream is big data — it never stops arriving (velocity), it mixes text, images and video (variety), and no single machine could hold or process all of it (volume).

Case file: Google's COVID-19 Community Mobility Report

Google's 2020 COVID-19 Community Mobility Report for New Zealand is a good real-world illustration of all three Vs at once:

• Volume — the report is built from location history data drawn from a very large population of device users across the country
• Velocity — Google needed to generate and publish the report within days of the national lockdown beginning on 25 March 2020, showing near-daily movement trends
• Variety — the report combines structured numeric drops (e.g. "retail and recreation: −91% compared to baseline"), time-series line graphs, and free-text explanatory statements about privacy methodology, all describing the same underlying dataset

(Source: 91908-res-2021.pdf, Resource E, and 91908-exm-2021.pdf, Question Three.)

Past-paper practice

What are the 3 V's of big data? Provide an example of each from the case study. (Adapted from 91908-cat-A-2023.pdf and 91908-cat-B-2023.pdf, Question Two (a), asked about TikTok and X respectively.)

Reveal how to answer well

• Define each V in a full sentence — don't just name it
• Pick one concrete detail from the given case study per V, not a generic example from memory
• If the case study doesn't obviously supply one V (e.g. veracity), say so and explain what evidence would show it — this demonstrates understanding rather than guessing

Applying this to an unfamiliar context

Activity

The exam will very likely give you a platform or organisation you haven't studied in class. Practise this transfer skill now: for any unfamiliar big-data scenario, ask yourself:

1. How much data, and how is that scale visible in the scenario? → Volume
2. What different formats or structures does the data come in? → Variety
3. How quickly does new data arrive, and does the use case need it processed immediately? → Velocity
4. How would you know if some of the data was wrong, fake, or unreliable? → Veracity
5. What would make this data actually useful, versus just large? → Value

Exit question

Pick a platform you use every day and explain, in your own words, why it counts as "big data" rather than just "a lot of data" — naming at least three Vs.

Key terms

• big data
• volume
• variety
• velocity
• veracity
• value

1. 1
2. 2
3. 3
4. 4
5. 5

← Back to Big Data overview

Next → Topic 2: Generation, Processing & Formats