Abstract
Video media has rapidly become the dominant form of internet content in terms of the amount of bandwidth that it consumes. It has also given rise to giant internet companies that deliver this content, such as Netflix and YouTube. I sought to examine how this content is delivered from two distinct views. In “Redesigning the Online Lecture Video Player” I look at video playback systems, while “Responsibly Regulating the Structure of the Internet” began as an examination of the systems required to bring content to users on the scale required by the current data-intensive video-dominated internet.
“Redesigning the Online Lecture Video Player” explores potential improvements to existing video players, with a focus on video players used for lecture videos. My primary goals with these improvements were to increase viewer engagement and improve the quality of feedback available to lecturers. This took the form of three distinct projects.
First, a more detailed logging system that expands on the basic video analytics offered to video posters by YouTube. This system tracks video views on a more individual (but anonymized) basis, allowing more customizability when displaying this data. This allowed me to not only replicate the graphs that YouTube provides, but also generate a plethora of tweaked versions which only display video re-watches or consider the speed at which the video was played, allowing video posters to more intimately understand how their videos are viewed.
Next, I integrated content quizzes with videos. Quizzes allow students to interact with lecture content and receive direct feedback tailed to their answers. These quizzes were displayed in the same space as the video, were accessible from the video at any time, and linked back to relevant sections from each of the questions in order to help students further review any questions they had missed.
Finally, I created a novel way for viewers to interact with videos in the form of video content searching. This system allows viewers to find a number of sections in the video that best match a text query that they input. This is achieved by searching a script generated through an audio-to-text processor and associating the best matching section of the script with a portion of the video. This system presents a novel method for interacting with videos, more akin to existing searching methods for text documents, which I believe would improve view engagement and overall experience.
In “Responsibly Regulating the Structure of the Internet”, I look the socio-technical system that governs interconnection agreements between suppliers and transporters of internet content. These agreements, when taken together, determine the structure of the entire internet. While this structure affects all forms of content sent over the internet, it is particularly pertinent for video content which both requires enormous amounts of data and is sensitive to the quality of connections available (particularly when streaming videos). It is no surprise then that some of the largest disputes have involved video providers like Netflix. I focus on the existing regulations that oversee how interconnection agreements are made alongside arguments for expanding or changing these regulations. I conclude that the current body of research and reasoning in the field is insufficient to create fully informed, responsible regulations for the ever-evolving field that is internet interconnections.
I examine two common arguments for imposing regulations on internet interconnections, as well as counterarguments and the apparent consensus, finding a heavy reliance on economic effects to justify policy decisions. Sets of internet rights and principles have been conceived, but there is little research into how to best achieve a system that supports these rights, or a thorough evaluation of how much importance the public places on these rights when compared with other concerns such as the economic potential of the internet. It may be that a maximally profitable internet and one which supports net neutrality are compatible, but neither this nor the public’s priorities have been sufficiently established.
I conclude that two key pieces of information are required before interconnection policies can be optimally regulated. These are the public’s desires for the internet and the effects of different regulations on non-economic factors. Without both, we are reduced to either trusting wholly in economics to bring about a desirable result, or regulating blindly with no knowledge of whether our intentions will bring about the results we seek.
Throughout “Redesigning the Online Lecture Video Player”, video playback features are scrutinized and supplemented, while “Responsibly Regulating the Structure of the Internet” examines the state of delivering content from provider to end user and proposes areas of research to improve the systems that govern this technology. Together, these projects shed light on the systems required to deliver video content to the end user, and provide a variety of proposals to further improve the user’s experience.
