I am working on creating a system that requires me to differentiate a email that is sent by a real person or a automated bot. I used gmail-api and got pretty sure that all emails under "Personal" label would have all the real life people emails, including important automated ones. Now how do I differentiate from there?
Asked
Active
Viewed 32 times
-1
-
Good news: if you manage to do that, you'll become rich and famous. Bad news: you probably won't. – JJJ Nov 04 '17 at 13:56
-
This is a research level problem. Yahoo research has published papers about this this year. And have now deployed it in Yahoo mail (possibly only their phone app), apparently. They did not just use headers though, they used the whole email. – Frames Catherine White Nov 04 '17 at 13:57
1 Answers
0
Tricky problem probably not solvable using the view of one mailbox content. If working across a large organisation where access and reading of all inbound email is implemented some sort of categorisation may be possible. Using the toolbox of spam filtering technology, Bayes filtering on keywords, urls, pictures etc. it might be possible to categorise email in to personal / work / commercial / bulk spam / transaction receipts. Check all other message headers as part of the email context.
Can you do the differentiate task by hand ? Automate how you did that.
Fred Gannett
- 111
- 1
- 4