A View from Emerging Technology from the arXiv
E-mailaholics: How to Tell If You've Got The Habit
A team at Yahoo Research work out how to classify individuals by their email habits.
For demographers and ethnographers, there can never be too many ways to classify individuals–by age, sex, ethnicity, work etc.
Now there’s another tyep of classification they can use to group individual: by their email habits developed.
Small worlds expert Duncan Watts at Yahoo Research in New York City and a few pals studied the time of day at which around 3000 individuals at a European university sent emails over an 83-day period as well as the email habits of over 122,000 e-mailers at a US university over a 2-year period.
They found two distinct types of emailer. They termed the first “day labourers” because they tended to send emails throughout the normal working day between 0900 and 1800 but not at other times. The second group they called “emailaholics” because these people sent emails throughout the waking hours from 0900 to 0100.
These groups appear pretty stable: roughly 75 % of users stay in the same group over a 2 year period. That gives a pretty good way of classifying individuals.
But there’s another beneft from the technique. Humans have a unique pattern of transmission that makes them easy to tell apart from machines that send spam. So the new method could be used to spot spambots too.
Ref: arxiv.org/abs/0905.0106 : Characterizing Individual Communication Patterns
AI is here.
Own what happens next at EmTech Digital 2019.