Intelligent Interfaces for Technology-Enhanced Learning

Abstract

When a user uses an image retrieval engine, he is in constant interaction with the engine, be it to create the query, provide feedback, view the results or evaluate engine performance. In this chapter, we have looked at most of these aspects. The retrieval engine must meet human needs. Therefore, it must be as close as possible to them. In particular, it must use the features that capture the semantic content of images, use similarity measures that resemble human judgment, have a user-friendly interface, etc. A lot of work has been done to that effect over the past few years; however, we believe much remains to be done.