The increase of bandwidth and streaming technology has made video on the Web the current "killer-app" of the Internet; however, users still face many problems.

This thesis investigates these problems, using surveys and user testing, and suggests ways for improvement. It researches the problems of indexing, browsing, retrieving and summarizing video content beyond the traditional metaphor of a fast-forward button.

Supplementary material like text is used to facilitate a better user experience with video on the Web. Users need representations of videos to make an informed assessment to decide if they want to watch a video and if so, what portions. Time-aligned text and visual descriptions can aid the user in making these decisions. A web-enabled prototype called mediacaptain emerged from the user studies and supports textual, graphical and visual descriptions. It is demonstrated on several content types and represents an advanced experience with video on the Web.