Using a collaborative filtering algorithm, Individual personalization serves up content recommendations unique to every single user based on their content preferences. We have developed intelligent formulae and filters that work along with the core algorithm, in order to deal with the ‘popularity bias’ problem that is typical to collaborative algorithms. We provide you levers to apply optimization parameters and also give you the option of selecting from 3 Algorithm variations. Analytics helps you track recommendation effectiveness and the effectiveness of your selections that are different from the default options. myreco suites of APIs talks to your carousel management systems so that you can track the effectiveness of the myreco carousels and also compare metrics with the carousels that are not powered by MyReco.
Fancy a ‘Trending in your area’ carousel that is completely automated and powered by AI? myreco can do that for you. Geotagging is common. We go several steps ahead and provide you the option to define a user segment by using any meta property or any usage metric. For example, would you like to serve a carousel that contains drama romance movies to the women (pardon the stereotype) and the same carousel that contains action movies to the men? Or for that matter, a carousel that serves the most-watched Hindi movies to all active Hindi users and the same carousel that contains the most-watched Tamil movies to all active Tamil users? All this is possible through the myreco mass personalization algorithms. Track effectiveness and compare with the non myreco carousels in Analytics.
Once you watch a movie, a set of similar movies appear as recommendations. myreco’s proprietary method not only ensures that the results of this carousel are superior to that of most other Apps, it also gives you the option to customize rules to suit your business requirement.
Each user gets several recommendation options that are personalized to his / her usage, which in turn are extracted from content meta mapping. Again, business rules ensure that results are most suitable to your business needs
Each personalization offering from myreco can be A/B tested before deployment. Analytics gives a day by day update on the test results and you can decide when to stop testing basis the results. End to end measurement ensures that post deployment, you can keep track of performance metrics.
When we receive inadequate or inaccurate metadata, myreco follows an enrichment process to deliver precise and accurate recommendations to users. The enriched metadata makes myreco more robust and helps to generate recommendations for different types of content-based personalization.