Over the last 10 years our knowledge base has become a pivotal feature of support, with a collection of digital services built around it. We get around a million searches and over 2 million article views a year, but search has its limits. Firstly, the obligation is with you to perform a search. And that's not all. You also have to:
- Choose the right keywords
- Assess results, and
- Refine your search again through trial and error
We can do better. What if we could recommend knowledge you (and only you) need without you having to delve and forage for it?
Our team has put together a prototype which does just that. We're using cognitive technology to select and propose articles that you've not read recently, but have been qualified as relevant, based on your article consumption history.
Before going further with the concept, we're making this first prototype available to customers to try out and share feedback. You can access your personalised recommendations here.
After viewing any article, a widget will appear for you to indicate if the recommendation was relevant or not, and there’s a field for you to share your assessment to help us fine tune the recommendations we serve up. We’ll read and evaluate every score and comment. The tool is just a prototype for now, so usual constraints apply: we can't guarantee availability or performance, English only and the design is basic.
