STACC’s recommendation system is an analytics solution, which helps to choose from a variety of news articles the ones that might interest the reader the most. It’s an artificial intelligence system that enables to:
- collect, store and process information about the online behavior of the readers;
- predict user intent (what the readers wants to do in the news site);
- predict articles that match the user profile (which articles are suitable for the reader at a particular time point);
- set business-specific rules (which articles can’t be recommended together, which articles are subject to special rules, etc.);
- conduct A/B testing (which recommendation model makes the best recommendations in terms of the business objective);
- monitor the recommendation system (how readers react to the recommendation system and how accurate the system is);
- improve the system constantly (using machine learning, the system learns to become more familiar with the reader over time and makes its recommendations more accurate).