What is data mining?
Definition
Data mining is analyzing large collections of data to find patterns, links and trends. Companies use it to predict behavior, target ads and score risk, and when the data is about people it raises privacy questions.
- Category
- Data
- Also called
- data mining meaning
- Last updated
NIST's glossary describes data mining as an analytical method for finding patterns in large datasets, with the aim of discovering data or knowledge that wasn't obvious. The term has nothing to do with the physical mining of anything: the data is the raw material and the patterns are what gets pulled out.
How it works
A company collects records, then runs software over them to look for groupings and relationships. A supermarket's loyalty-card data might show that people who buy one product often buy another. A bank might look for transaction patterns that match fraud. The same methods applied to personal data can predict what you'll buy, whether you'll repay a loan or what you'll click.
Much of the data comes from you: purchases, location, searches and app activity. Data brokers add to it by buying, joining and reselling records, so a profile can include things you never told the company directly.
When it becomes a privacy problem
Mining by itself is neutral. Trouble starts when the data is about identifiable people. Patterns can reveal things you didn't share, such as health conditions or finances, and the results can be used to target or exclude you. Even data stripped of names can sometimes be traced back to a person when combined with other sources. NIST also lists privacy-preserving data mining, a version that works on data altered to hide sensitive details.
Data mining vs. related terms
Data mining finds patterns in data. Data aggregation is the compiling step that comes before it. Web scraping is a way of collecting the data in the first place.
Why it matters to you
Data mining is how scattered scraps of information become a profile. Less data in circulation means less to mine.
How to protect yourself
- 1
Limit what you hand over
Skip optional questions on forms and loyalty sign-ups, and use a separate email address for them.
- 2
Turn off ad tracking
Switch off ad personalization in your phone's privacy settings and in your Google and social media accounts. Reject non-essential cookies.
- 3
Get your name out of broker databases
Data brokers sell the profiles that get mined. Serus finds where your address and phone number are listed on data brokers and people-search sites and removes them for you. Our data broker guides cover each site.
- 4
Use your access and opt-out rights
Where you live, laws such as the GDPR and California's CCPA let you ask what a company holds on you and ask it to stop selling or sharing it.
See where your personal data is exposed
Serus finds it on data brokers and people-search sites and removes it.
Frequently asked
What is data mining in simple terms?
Using software to find patterns in large amounts of data, for example which products people buy together or which transactions look like fraud.
Is data mining legal?
Generally yes, but when it uses personal data it has to follow privacy laws such as the GDPR, which put limits on how data is collected and used.
Is data mining the same as data scraping?
No. Scraping collects data, usually from websites. Mining analyzes data you already have.
Can data mining identify me?
It can, especially when datasets are combined. Removing names doesn't always make data anonymous.
Sources
Something wrong or out of date? Tell us and we'll re-check it.