Here at Kinonation, we have a fast-growing catalog of titles on our platform, accessed by many streaming services around the world. From early on in our company we have been working on an objective rating system for our titles — we call it our Kinoscore. Our early algorithm worked reasonably well when we had a small number of titles in our catalog, but with our rapidly expanding catalog, it has been showing some limitations.
Now that we have over 12,000 titles, it is time to update and revise the methodology used to calculate Kinoscore. Leveraging our growing big data set, we now use advanced machine learning and AI to develop a more robust algorithm. We now also include performance data to provide both an objective measure for the Kinoscore as well as the additional data with which to refine the accuracy. With that we’re on track for meaningful Predictive Analytics, driving video content matching and demand forecast more with data, and less with guesswork and “gut feel”.
We are now starting some real-world tests of Kinoscore with select streaming services. The new Kinoscore should enable us to provide our streaming service partners with better recommendations for their curated title orders.