How's it going to be able to identify anything without any of the necessary information?
How's it going to be able to identify anything without any of the necessary information?
theHERETICS - Brute Force - Sonata - Dreams - The Pulsing Trollfags - The Expendables
Visit my home for banned, neglected, and otherwise disenfranchised players on Discord!
Well, the user data wouldn't be disclosed to developers of course, but even if it was it's hopefully hashed and salted. The git project would be about the algorithm to detect and track user patterns, there is no need to have actual user data for that. Just a couple of demo accounts is enough. I get the feeling you don't know much about his subject?
Oh you just want people to imagine solutions as opposed to using analysis. I see now!
theHERETICS - Brute Force - Sonata - Dreams - The Pulsing Trollfags - The Expendables
Visit my home for banned, neglected, and otherwise disenfranchised players on Discord!
No, Mr. Troll, the data comes AFTER the algorithm is implemented. First you build the tool to track user behaviour, then the tool collects user data, THEN you analyse said data. Does that make sense?
Also, you are clearly not a developer or have any expertice on the topic so I don't know why you're here.
I am not a developer! Was considering it from a machine learning perspective under the assumption that they already have a backlog of great data :(
If you're collecting the data anyway what's the difference exactly as far as who has access to what?
theHERETICS - Brute Force - Sonata - Dreams - The Pulsing Trollfags - The Expendables
Visit my home for banned, neglected, and otherwise disenfranchised players on Discord!
I have no idea what kind of backlog of data already exists, I'm just a lowly player with no dev insight :)
The point of collecting user data is to assess user behaviour. Once there is is enough data, an algorithm should be able to determine what is unusual user behaviour. Say for example the 6 hypothetical work mates always log on with an interval of 5 minutes - that's probably one guy. But if they are spread out throughout the day + all go to different IPs after work hours they're probably not one guy.
These are just low level examples of how machine learning and pattern recognition works. Once the data is there you could of course couple it with time zones, devices etc to create a truer picture of any user. That would create a much richer picture of how a user plays the game than just going "Same IP detected - you are barred".
@chadde: The point of collecting user data is to assess user behavior.
Sounds like EDGE, Firefox, Google all rolled into one. 1984, Big Brother. Wanna collect my data.
@BigBrother: if logs are spread out throughout the day, + all go to different IPs after work hours they're probably not one guy.
Would not work so well for college campus dorm dwellers who all use same IP day and night
Phone with unlimited data, cost more but if ya love the game.
There are currently 1 users browsing this thread. (0 members and 1 guests)