A
safer and
faster approach than "predictive domaining" is probably "retrospective domaining".
Learning from the past, from the mistakes of others, from experience of others, from the fails and successes of others, etc.
Even if you want to be a successful predictive domainer, you need to be a retrospective domainer with some experience and feeling for the whole domain market first. You can't just start with domaining, go to google news, type 'future trends', then register some word combinations you read in some articles and wait for $$$$$$ and call yourself trend domainer.
Last but not least, you have to be lucky. Many apparently new promising trends, can turn in to fails because of different circumstances and developments.
http://venturebeat.com/2013/07/05/bbc-is-done-with-3d-tv-after-the-doctor-who-special/
http://blog.sfgate.com/techchron/2013/07/08/your-ability-to-focus-why-3d-television-cant-take-off/
Then again, this could change with new technologies and development.
No one really knows.
As i said, you need some luck too.
I personally like trends and also participate in trends, hypes and whatever you want to call it from while to while, but i know that most of the time its gambling and just a game with unknown outcome. so its a rather small part and i try to diversify as much as possible.
Its fun though, no doubt.