I make no claim at being an excellent interviewer. Masters like Frost, Jennings, Sawyers, Woodward, and Wallace set the bar for being able to craft text and subject to provide profound stories that leaves even the most ambitious wordsmith trembling.
However in most situations, working sans journalist or writer, the shooter is forced to serve as both visual and content artist. This is not as challenging as it sounds but it does require a bit of careful work.
If the subject isn't ready or perhaps caught off guard, try and give them a bit of warning (Though sometimes this forces them to try beat a hasty retreat from the lens). This plants the seed in their head and hopefully has them prepping the gist of what they'd like to say in mind. As an interviewer, the aim is answer the foundation of informative content: The W's. "Who, what, when, where, why, and (w)how," are not just for writing! Whether its journalist content or commercial, these are usually key deliverables. Consider the example:
Who: Susan Lakin, RIT School of Photo, David Halbstein, and Majorie Searl
What: Student Work Exhibition
Where: Memorial Art Gallery
How: Augmented Reality Platform
These questions are answered in the first 30 seconds. "When" is answered more in terms of the collateral information on the website (on a YouTube Page) "Why" is posed more by the context where it is distributed (by the School of Photo as well as the Memorial Art Gallery). I prefer to keep the date out. The inclusion specific dates seem to sound a bit canned and formulaic, causing issues down the line if the content is redistributed. However, it may work for work that is self standing, or focused on reporting a specific event rather than telling a story.
Specifically to DSLR work though, clip length becomes an issue. With tape based, even other digital formats (read, real video cameras) the length of a clip is limited by either battery or storage space.
Working with a DSLR, because of technical limitations, clips are limited for 15-20 minutes before a timer in the camera shuts down. Due mostly to heating/cooling issues with the sensor or a memory buffer limitation, nothing is worse than missing an integral part of the story because you have to restart the camera. Though this isn't a foreign concept (tape/film/card changes have always been a factor) going from a 45-60 minute cut to 15-20 puts a real kink in the workflow, and keeping an eye on the clip length while trying to maintain rapport with the subject just compounds the issue on top of monitoring light, audio quality, and keeping the next question in mind.
That being said, with a bit of practice and good peripheral vision, a one-man interview/shooter isn't the worst. Make sure the sound is solid, and you answer the W's and one isn't too far away from a solid story.