29th Annual Student Research Conference
A Graph Theory-Based Approach to Product Recommendation using Functional Programming
John M. McKeever
Dr. John Seiffertt, Faculty Mentor
User to user connections are an increasingly important tool quickly becoming standard practice for product recommendation. By warehousing book data and comparing the preferences of readers, we can write computer algorithms to generate book recommendations for individuals. Our approach uses functional programming, in which functions are evaluated mathematically. The algorithm developed for this project involves a breadth first search on a graph of reader nodes connected to book nodes. Reader nodes have connections to book nodes, representing the reader's opinion of a particular book; readers connected to the same book define a path by which books may be recommended to both users. Functional programming is particularly well suited for this task, as it allows for high levels of parallelism, enabling the system to scale as the database grows larger. Results thus far have indicated that the algorithm returns appropriate recommendations based on the readers' preferences.
Keywords: Computer Science, Functional Programming, Graph Search, Product Recommendation, Book Recommendation
Topic(s):Computer Science
Presentation Type: Oral Paper
Session: -2
Location: VH 1328
Time: 9:45