Social Recommendation

Strands technology is based on human-generated links between items (songs, videos, products, etc) much the way Google uses human-generated links between web-pages. Complex networks of human-generated links between items are the best mechanisms to analyze how our society considers and uses the products it creates and consumes. Although the basic principles used to construct the networks are simple to understand, the sheer amount of data makes the networks themselves huge. These networks, which are the foundation of the Strands recommendation engine, provide great flexibility and power.

A critical focus of our recommendation technologies concentrates on organizing observations from people's behaviors and distilling that information in a way that can be leveraged into a useful purpose, representing knowledge of behavior as data and making it possible for electronic devices (computers, mobile phones and other Internet-connected devices) to access that knowledge.

When it comes to delivering personalized recommendations to individuals or groups, Strands' recommender engine considers that taste is context-sensitive and evolves over time. Strands provides real-time recommendations from the first user interaction without the need to ask the user to rate content or build a profile. Our technology can deliver relevant suggestions to first-time anonymous users as well as providing filtered, personalized results based on individual profiles.

There are three Key Design Drivers of this technology: Reliability, Scalability, and Fast Response Time.

Some of the characteristics that make Strands' Recommender System unique are:

• Content agnosticism
• Platform agnosticism
• Scalable catalog and query support
• Efficiently extensible recommendation strategies
• Simple, elegant and robust for large scale deployments
• Delivery of Instant recommendations with no training period