- IRIS.TV’s algorithm uses historical engagement data (User-Asset Data), group and individual preferences (User-User Data), and asset similarities (Asset-to-Asset Data) in addition to its proprietary recommendation technology.
- All requests to the IRIS.TV API require a UserID to remember user preferences and generate personalized recommendations. In IRIS.TV plugins, this UserID is an anonymous character string that is then stored on a cookie, which will expire after 15 days.
- In custom integrations, IRIS.TV recommends a similar character string. The publisher can associate this string to a login or device in order to ensure personalized recommendations for logged in users, but this is at the discretion of the publisher.
- UserIDs are generated via two methods: Publisher-provided loginIDs and dynamically-generated cookies. Please see the diagram below for details.