Content Recommendations Uses Psychological Preferences
A user behavior modelling algorithm has been developed that acknowledges a user’s changing preferences on sites where users experience many pieces of successive media and adjusts the recommendations accordingly. How much time does it take to bore a customer with the same “reliable” recommendations? How long will it take for a certain recommendation to finally pique the interest of a user? These are questions that are answered by utilizing algorithms based on user psychological preference state. This system uses latent preference states “Sensitization”, “Boredom”, and “Recurrence” to predict the average user’s preference evolutions, maximizing the potential for conversion purchases and visitor satisfaction.
Media sites such as Spotify, Pandora, and Youtube and e-commerce sites like Amazon and eBay utilize media recommender systems to serve their visitors. Recommendation algorithms that do not take into account users’ changing preferences struggle in situations where the user is repeatedly consuming the product or media. Repeated exposure to the same recommendations or media result in uninterested consumer, and immobile recommendations can bore a user after some time.
BENEFITS AND FEATURES OF RECOMMENDATION ALGORITHM USING PSYCHOLOGICAL PREFERENCE STATE:
- Utilizes proven psychological patterns to determine user preferences over a long visit
- Avoids repeated unwanted recommendations that annoy and deter visitors
- Promotes visitor conversion by accurately predicting preferences
- Optimizes continuous media experience and purchase vs. 1-time purchases
Phase of Development Prototype developed and tested