The Rise of Video in Task-based Documentation

By Sean Healy

The adoption rate of DITA has been truly remarkable. While its growth stems from and is most apparent across those organizations that publish technical documentation, DITA is certainly not limited to technical publication departments. Because of its specialization feature, DITA is now viewed by a variety of businesses as a stable, low-cost solution to all sorts of documentation from marketing to financial reports to training material and so on. By some estimates, up to 80 percent of new XML publishing implementations will be DITA-based by the middle of 2009.

Also on the rise is the use of video deployed to the Web, which should be of no surprise to even the occasional Internet user. Anecdotal evidence of this ascendancy has to include YouTube, the fastest growing site in Internet history. On average, 100 million videos are streamed from YouTube every day. 65,000 new video clips are uploaded every day. 13 million viewers visit YouTube every month. One has to look no further than the home page of the United States White House website (http://www.whitehouse.gov/) to see how mainstream Web video has become and how much value has been placed in its ability to communicate a message to a broad audience.

So why haven't we seen more video integrations within structured documentation implementations such as DITA? At face value, video would seem to be a perfect fit with task-based procedural documentation because of its ability to show how to do a task or series of tasks.

The lack of such implementations is due in part because video is largely a black box. The information within it is rarely described with the necessary level of precision. Most viewers will not watch 10 minutes of video to find 30 seconds of relevant material. Many how-to video companies (http://mashable.com/2007/05/14/video-howtos/) provide users with search capabilities; however, the level of granularity for these videos is usually too high to be applied to technical, task-based documentation where the ability to quickly find and understand discrete steps is key. Video chunked into relatively large pieces by typical methods is as unwieldy as coarsely-chunked unstructured documentation. Content reuse and searchability become at best difficult or at worse impossible.

What if steps within a DITA-based topic had video links to show how to complete a discrete step or procedure? We have all used search engines that allow us to enter criteria that return topics where the text is found. What if we entered search terms that take us to a particular video segment that demonstrates how to complete a specific task? What if we could browse clickable thumbnails to find relevant video segments? What if every time a video segment displays, its associated text-based topic displays as well?

MPEG-7 (http://downloads.wildbasinmedia.com/MPEG7_DITA_1.pdf) is an important component in the advancement of video-integrated documentation, and provides a bridge to these solutions. MPEG-7 is a multimedia (video, audio, voice, images, graphs, 3-D models, etc.) content description standard. Its tagged data can be passed into, or accessed by, devices or computer code. This description standard is not aimed at any one application in particular; rather, it supports a range of applications. MPEG-7 uses XML to store metadata that references multimedia instances. It provides fast and efficient searching, filtering, and content identification when coupled with XSLT or XQuery. MPEG-7 is not to be confused with other standards developed by the Moving Picture Experts Group (MPEG) such as MPEG-2 or MPEG-4, as these deal with the actual encoding of moving pictures. MPEG-7 makes it possible to separate the information about video (i.e., metadata) from the video itself. This separation of concerns provides the key to the multimedia search conundrum, using a standard XML toolset and workflow.

Elements of the MPEG-7 schema can be attached to video time-code segments in order to tag particular events within video such as particular tasks within a topic. These video segments are then retrieved and viewed through the corresponding metadata. Video components described by MPEG-7 can be integrated into online documentation and made accessible to users in a similar way as are, for example, text-based components in DITA. From an author's perspective, adding video to a documentation set is as easy as inserting a resource id into the XML source -- just like one would for an image. Like their text-based counterparts, video components may be reused across topics and documentation sets. MPEG-7 offers a standardized approach to the management of video metadata.

For the greatest user experience possible, video production must adhere to quality standards and best practices. For example, even though video segments may be remixed across multiple documents, captured by multiple video teams over time, and processed by more than one editor, for users video segments should all appear as if they have originated at the same time and in the same place. Technical publications departments deal with disparate writing styles among their writing staff by using a style guide. Similarly, the medium of technical video requires its own style and best practices guides to govern lighting, shot angles, sound quality, video formats, common workflow processes, etc. Like text-based content and images, video segments are subject to revisions and updates. Fortunately, MPEG-7 also provides a systematic way of dealing with video revisions, too.

To be sure, there are special considerations when using video in technical documentation, but imagine the benefits of having a methodical way of viewing discrete steps within, for example, an airplane engine maintenance manual. Mechanics could search across MPEG-7 repositories to find and view appropriate video segments and read the associated DITA-based documentation adjacent it for further reference. Because many parts and procedures are identical across plane models, standard reuse and effectivity principles can be employed for video in the same way as for text-based content. Just as task-based documentation is modularized so that new documents can be created from pieces of existing text, video segments, each assigned a unique id, can be reordered into new configurations. In short, the same themes (e.g., reuse, modularization, metadata employed for increased searchability, reduction of information redundancy) that run through the design of structured documentation can be applied to video with the help of MPEG-7 and associated technologies.

To this point, I've only mentioned camera-generated video as benefiting from an MPEG-7 implementations when applied to structured documentation. However, software-generated video (i.e., screen capture software such as Camtasia, Jing, Captivate, and IShowU) can be segmented, managed, and deployed in the same way. For example, it is possible to repurpose your existing walkthrough video of your latest software product, so that video segments can be searched and viewed. Because each segment is allocated a unique id and associated with metadata, segments can be reordered and reused where appropriate. This type of video is not equipment or software intensive. In fact, you may already have almost everything you need already.

With new video standards and practices, it is now possible to merge video seamlessly into structured documentation for the Web with greater control over its management, search and presentation. The convergence of greater bandwidth speeds, improvements in production tools, and cost-cuts in video production workflows make online video-integrated documentation (OVID) a viable enhancement to your structured documentation – technical or otherwise. It is conceivable that in the near future, video will be an integral part of structured documentation, just as it has in the Web at large. For an example of video integration, see http://screencast.com/t/z0OyhgXn6cV (camera-generated video) and http://screencast.com/t/gQpxl6Fiz (software-generated video).

About the author:

Sean Healy is owner of Wild Basin Media (http://wildbasinmedia.com/), a company that specializes in blending multimedia with structured documentation. He has worked as a software architect, software developer, XML consultant, technical writer, and project manager. He holds an M.S in Technical Communication from the University of Colorado.