The State of Education Video 2017

Activity Reporting

Modern video platforms typically provide dashboard-style activity reporting, but what is needed is for the video service to be able to report activity back to another component serving as a global event aggregator. In a modern LMS, practically every click a user performs is logged, and those logs are available for administrators to inspect. Typically, these are used to help a student who is having technical problems using the course website: The teacher or other support people can check the activity logs for that student to see exactly what was done and therefore easily provide guidance to avoid confusion moving forward. There’s also great research value in these activity records, where we might be able to mine the data for insights on improving how we teach. Anything happening outside the scope of the LMS’s event listening range won’t be recorded. This is currently a limitation with LTI: Once someone starts working in the LTI, the LMS knows nothing about the activity aside from any updates to the gradebook entry representing it. It would be preferable for raw logs to pipe back and be aggregated with the rest of the activity logs, rather than exposing the logs only in a digest form as a dashboard within the component’s user interface.

Versioning

The service manager for an NGDLE would want to periodically take whole system backups or snapshots, and it’s extremely hard to imagine how that would happen without some sort of command to all components in the system to snapshot themselves at that time. In this case, a well-maintained NGDLE will likely be more easily maintained than a monolithic LMS, plus whatever isolated add-ons are in use, since the components of an NGDLE should be strictly decoupled and only interact using standardized interfaces. If something goes haywire, it will be easier to rule out an unintended consequence from an undocumented interaction between components (since they ought not to exist). And since the system is more distributed, it’s less brittle to catastrophic hardware failure. Still, some systematic sysadmin “undo button” feature will be needed. For most video platforms serving educational video, this wouldn’t be a horribly expensive feature addition since educational video has a generally low turnover rate compared with, say, security camera systems.

Data Ingest/Export

The NGDLE paradigm is attractive because it offers the promise of adaptability and freedom from vender lock-in. That attraction is a mirage if the content isn’t reasonably portable from one service to another. In the case of video, this is a significant challenge, since we handle very large amounts of efficiently compressed data. The ideal scenario would be one where a school could migrate its data from one video service to another as simply as it could provide the new platform with temporary credentials on the old platform and allow it to copy over. Less convenient yet standards-based and wholly tolerable alternatives could be imagined. It may seem unlikely that a vender would invest in such a capability, but this is a case of the door that swings both ways. If a potential product can both ingest your existing catalog of content and provide you with guarantees that you’ll be able to get it out again, it’s a far more attractive product than without those assurances.

Personalization

One of the goals of the NGDLE movement is tailoring the learning environment to the user: so-called adaptive learning, where the way students engage the content is somehow customized based on previous experience with the student, and knowing what works best for each of them. This sounds terrific, but there are serious implementation problems to making it work in the kind of distributed confederation of services model being advocated while assuring privacy requirements are met. To put it bluntly: How does a system effectively communicate to all of its components personalization-relevant information such as a particular student who is hopelessly lost on trigonometry, and of what value is that information to a video platform? Many of the components a student would engage with in the NGDLE would be third party, so privacy issues when it comes to sensitive data of this sort are consequential. Most likely, the video platform would not be informed, and a remedial video would be added to the course content using logic at a higher level than the video platform operates. It will be interesting, should the NGDLE revolution erupt, to see how this tension between delegation of sensitive information and distributed, nonhierarchical architecture plays out.

Here Be Dragons

The educational technology market is a somewhat small pond. In this market, Blackboard is a big vendor, and it has annual revenues estimated at about $1 billion. Even the unimaginably huge players, like Pearson or McGraw-Hill, have relatively low revenues.

Out in the greater world, there’s a company like Apple, which takes in much more money each quarter by an order of magnitude than a very successful video platform vender like Kaltura does. That comparison would be completely unfair were it not for the fact that Apple is relaunching iTunes U as a full-featured learning management system, instead of merely a media delivery platform.

Similarly, with 2016 revenues of $85.3 billion, Microsoft offers the Office 365 cloud platform. You might think Office 365 is merely a subscription version of the Microsoft Office suite in the Adobe Creative Cloud mold. But one of the cloud applications available in the Office 365 Education products is the corporate video portal.

Inside this portal, a user may create a channel, lock it down to a set of people (for example, students enrolled in a class), upload her course video to it, add captions for multiple languages, and embed the videos in her course website. The video streams to HTML5 video elements at adaptive bitrates using either HLS or MPEG-DASH from Microsoft Azure media servers, depending on browser capabilities. A reporting dashboard is provided for each video, showing a graph of when the video has been watched, plus an engagement graph showing the percentage of viewers who watched at segments of the video: whether viewership falls off at a certain point or students tend to skip ahead to a particular point.

These are state-of-the-art features for delivering educational video. Microsoft also offers a free plug-in for PowerPoint called Mix. It provides a tool for recording videos of your presentation that allows for uploading a video to one of your Office 365 channels with a few button presses. In addition, Microsoft also offers its own learning management system called Microsoft Classroom, plus School Data Sync, a tool for binding the Office 365 environment to your school’s Student Information Service, an enterprise IT system that manages final grades, transcripts, and current class rosters, among other things. And it’s worth mentioning that the Bill & Melinda Gates Foundation had partnered with EDUCAUSE on the research leading up to the NGDLE white paper.

On a slightly less disruptive note, Instructure is the company behind Canvas, the LMS with the fastest growing market share over the past several years. Instructure was an early adopter of the LTI standard, and it remains a reliable supporter of LTI by running EduAppCenter.com, a collection of LTI applications that teachers from K–12 and beyond can add to their courses. Instructure recently deployed its own video platform called Arc that promises novel features for encouraging student engagement.

So we see two tech titans committing to the educational technology market, plus the ascendant LMS vender diversifying into the video platform area. Again, these are the sorts of developments one expects to see when a product sector matures.

These are truly exciting times for educational video.

This article appears in the March 2017 issue of Streaming Media magazine.

Free

for qualified subscribers

Subscribe Now Current Issue Past Issues