Semantic networks are often used as a method of relating multiple pieces of data to each other. ConceptNet is a semantic network that contains information about words and how they relate to other words. ConceptNet and other semantic networks are often hosted remotely and accessed as a service, and data retrieval times can be large. This project examines multiple data caching strategies and their impact on the performance of two existing applications that make use of ConceptNet data. We found that the largest factor in whether or not caching improves the performance of semantic network applications is the access pattern of the particular application.
Worcester Polytechnic Institute
Major Qualifying Project
All authors have granted to WPI a nonexclusive royalty-free license to distribute copies of the work, subject to other agreements. Copyright is held by the author or authors, with all rights reserved, unless otherwise noted.