Improving usage pattern quality by comparing different sequential pattern mining methods and the effect of considering additional user information

Frequent usage patterns generated can provide valuable information for several applications such as platform restructuring and recommendation. In this project, we aim to compare different practical methods, and to investigate the effect of user identity and user intention information on them. To that end, a technique and a framework need to be developed, in which frequent patterns are composed of more refined analysis result instead of simple frequent sequences of basic operations over all users behavior. The outcome of this project is expected to improve the user experience for the partner organizations product and such methods can be also used in various relevant applications.

Faculty Supervisor:

Fred Popowich

Student:

Zelin Tian

Partner:

Kinematicsoup Technologies Inc

Discipline:

Computer science

Sector:

Media and communications

University:

Simon Fraser University