Telco traffic classification

  • Case Details
    Is it possible today to imagine our everyday life without personal mobile phones? Using phones for everyday communication, telemarketing purposes, verifying identity through SMS/Voice, getting information about a product from call centers… there are plenty of telco traffic cases with a large growing potential. Automatic identification of telecommunication traffic classes can have a great benefit for both business and safety of end-users.

    Every month billions of transactions are executed, and the goal of this task is to classify international telecommunication traffic. The classes are not defined in advance and your task is to distinguish between various behaviors: legit or fraudulent user, call center or application number…

    For this purpose, we have prepared a sample of international and roaming telco traffic with all necessary information: hashed calling/called number, voice call duration, operators’ names/countries, phone types, etc. You can use various machine learning and statistical techniques and tools that can help you in pattern detection.

    Join us and let’s play with data!

    TeleSign Data Science team

How to apply ?

You can apply either as an individual or as a team.
Teams need to consist from 3 to 4 members.

We have limited number of places, so selection process will occur. You shall be notified about the results by April 1st