Decoding Content with Machine Learning

Ever wish you could decode what really piques a user’s interest in their online journey? Or how people speak in one space online over another? Then you’ve come to the right place – as Dublin-based Kevin Koidl has been hard at work realising this ambition for the past number of years.

Converting Research into an Enterprise

First inspired to investigate the world of content on a deeper level as he researched his PhD, Kevin has turned this interest into an enterprise. The co-founder of cognitive content platform, wripl, he works alongside a team to apply machine learning to identify trending content topics, compare competitor sites and determine the best way to create and deliver relevant content to target audiences.

An exciting project, Kevin says: “It was originally a content recommender system. At the time, we created user profiles based on reading behaviours that would trigger that next click.”

“But now, we’ve built a cognitive content technology around content that is primarily text-based, like articles, dialogue streams, images with alt text or video captures. And we’ve developed a cognitive part to allow clients to adapt content to certain audiences or contexts.”

With a roster of impressive clients and his sights set on bringing the platform to new audiences, Kevin adds: “In a nutshell, you have a platform that allows you to understand what language your audience use and in what context, which helps your content production. And then we can help on the content delivery side as well, to ensure content reaches the right audience at the right time.”

We’ve built a cognitive content technology around content that is primarily text-based, like articles, dialogue streams, images with alt text or video captures. And we’ve developed a cognitive part to allow clients to adapt content to certain audiences or contexts.

Decoding Language

With increasing competition for audience attention online, the platform offers partners and brands the chance to speak to their target audience in their own language. And it’s all thanks to technology.

Kevin explains: “To give you an example in practice, let’s say you want to sell a book about the foodie scene in Dublin. And the foodie scene has a certain vocabulary and way of discussing new books. To uncover this, we can mine the terminology that community is using. We pull out this vocab from user reviews on Goodreads, mine this then to identify the semantically rich keywords, and then use this to inform the language you use to sell the book.”

Formerly based in Germany where he gained valuable experience in the automotive industry, the tech community in Dublin is a natural fit for Kevin. He says: “We’re a small innovative company so we have a core team of four to five full staff, and depending on demand we can then call on this network for support. And Dublin has a great network of people who are experts in this tech and research space.”

We’re a small innovative company so we have a core team of four to five full staff, and depending on demand we can then call on this network for support. And Dublin has a great network of people who are experts in this tech and research space.

Sharing Insight with the Tech Community

Along with working with a host of partners and a wider network, Kevin and the wripl team are also sharing their insight into the application of machine learning for commercial purposes with the wider tech community.

“We’ve started cognitive content strategy events. We ran one two weeks ago, and it was great. It was a mix of people from the start-up world interested in machine learning and then a corporate audience interested in looking at how this could apply to their companies or objectives,” Kevin reveals.

With the inaugural event taking place in Bank of Ireland, Kevin is hard at work planning the next meet-up in April. Stay tuned to our events page to find out more. And check out Kevin’s work on wripl and his thoughts on AI versus Machine Learning on LinkedIn.