The IPTV dataset contains the following features with more than 5 million rows and the size of data is more than 1 Gigabytes.

• User ID
• SESSION START (DATE/TIME)
• SESSION DURATION
• PROGRAM NAME
• PROGRAM DESCRIPTION
• PREOGRAM GENRE
• PROGRAM CLASS
• SERIES TITLE
• SERIES NAME
• Episode title
• Episode number

1. Technical (algorithm) challenges ( 80% of total score )
   • Build machine-learning model to predict the top 10 items (movie, series) which the user will watch in future (60 %)
     • The final model should take as input a list of (user IDs) and output the result for each user
     • The list of items (movie, series) to choose from (rank) will be given
     • The user that we will test on, will have historical data in the training set
   • Determine user profiling (20 %)
     • How many real users for the same account and their gender and age
2. Presentation challenges ( 20% of total score )
   • Determine the best way to present the result (i.e. long tail plot for a given user)