Intelligent Recommendation System is developed as a Service with multi-tenancy design. Incorporating domain-knowledge can be extended to many cross-domain applications, such as multimedia recommendations, online advertising recommendations, CRM or decision support... to meet the needs of the industry.
Intelligent Recommendation System is developed as a Service with multi-tenancy design. EC vendors could request recommendation services. System includes Data Ingestion, Batch modeling, Nearline storage, Recommendation API, Evaluation, A/B Testing, Management Console modules.
This two-stage composite recommendation technology integrated multiple analytic methods and online usage context detector. Besides, A/B Testing mechanism has been established to simultaneously compare and test a variety of recommended algorithms performance through streaming technology based on optimizing the effectiveness of the recommendation technology.