top of page
A Personalized Activity Recommendation Web Service (Java)
-
Developed an interactive web page (HTML5/CSS/Javascript) providing 3 main APIs to handle the HTTP requests of updating preference, searching activity and pushing recommendation.
-
Built a customized web service (Java servlet, REST API) to fetch restaurant data from TicketMaster API and store user preferences (MySQL/MongoDB).
-
Improved content-based recommendation precision/recall by developing a collaborative filtering algorithm and matching similar restaurants based on categories.
-
Deployed server side to Amazon (Cloud computing) EC2 which is able to handle 150 queries per second (QPS) tested by Apache JMeter.
Try me!
bottom of page