User and content data aware recommender system
Recommender system making smooth coldstart
Predicting user content recommendation was never easier. Our recommender system can work from cold start by using content information. It carefully catches every user activity to improve recommendation quality during initial user data collection.
System functionality
  • Recommendations from empty user data bank (content only)
  • Recommendations based on user behavior that uses typical user actions for similar events
  • Rapid collection of user behavior information, utilizes every droplet of user activity
  • Building user content awarness map for heuristic building
System application examples*
* NDA issue:
These examples are not real cases. Cases of our clients are all under strict NDA and we cannot demonstrate them, these examples are solutions that can be build with our components
Cold start
Allows working from small or zero amount of user behaviour data
Reason of every recommendation is accessible and can be expanded more according to your request.
Can be configured for your recommendation task and domain knowledge of your experts can be fully utilized
Easy to install
Compatible with multiple systems including hierarchical ones
Component features
Can run on limited equipment
Modularity and full customization
Outputs a rich data stream
Browser visual interface
Builds user awareness set
Robust (processing speed up to 0.01 sec.)
API infrastructure integration
For further information feel free to contact us