News for hackers. Stuff that matters.
Re:Database containing details of 56M US residents found on ... by swarnie_ on Friday January 10, 2020 @11:02AM (#22011868)
General question to HN because i don't know the answer.

Has anyone ever tried de-duplicating and matching up records between multiple leaks to build user/person profiles?

I guess with enough unique keys, compute power and time you could build a reasonably accurate profile of a person by matching up email addresses, phone number or SN numbers?

Would probably make identify thief and fraud a lot easier going forwards.

Re:Database containing details of 56M US residents found on ... by astura on Friday January 10, 2020 @02:16PM (#22014240)
Data broker-type companies do this already and sell it. They even include stuff like your probable religion, probable income/net worth, probable hobbies/interests, etc.

For example:

>Each record contains entries that go far beyond contact information and public records to include more than 400 variables on a vast range of specific characteristics: whether the person smokes, their religion, whether they have dogs or cats, and interests as varied as scuba diving and plus-size apparel.

https://www.wired.com/story/exactis-database-leak-340-millio...

HN Discussion: https://news.ycombinator.com/item?id=17421140

Re:Database containing details of 56M US residents found on ... by dannyw on Friday January 10, 2020 @11:30AM (#22012126)
You have, quite literally, described Facebook.
Re:Database containing details of 56M US residents found on ... by anon73044 on Friday January 10, 2020 @12:12PM (#22012518)
Probably not much more than what was in the Equifax breach.

I think one of the bigger issues in generating profiles on ~325 million people would be merging tons of incomplete data and lots of old/outdated data. Not to say that you couldn't use some machine learning to fill in the blanks, but I'm sure most if the big data/analytics companies are already doing that.