Visitor #00001, we have been expecting you.

This is your profile, created by FROSMO based on your browsing behavior. Each time you interact with the website, we collect data to create a relevantly personalized experience for you.

Introducing Natural Language Processing – The way to personalize content

Tom de Raad

Natural Language Processing, NLP in short, is all the rage — and with good reason. With this AI solution, computers can accurately read and understand human language, which means that computers can also reply in kind. Combine that with a computer’s infinitely scalable memory and instant response time and serving humans with exactly what they’re looking for just got a whole lot easier.

NLP is used as a spellchecker in word processors. Some email systems use it to suggest canned answers based on your writing style. Search functions can use NLP to suggest terms and come up with results when your original query doesn’t produce any. The better translation software does so with NLP. AI chatbots analyze questions in the same way to provide accurate answers or further reading material. And the best thing? Every input and every suggestion provides a new data point and learning opportunity for the algorithm.

What makes NLP so interesting for Frosmo is its ability to accurately suggest similar relevant content, which is perfect for personalization.

The many business benefits of NLP

For a major university we do just that; suggest further reading material by matching articles based on their similarities. Frosmo uses a neural network-based NLP algorithm that analyzes all their articles by preprocessing them and converting the text, titles, ingress, and so on, into vectors, numerical representation of the article. After which the vectors are used to score the articles for similarity and rank them in order of relevance. 

The benefits for the reader are obvious as it introduces them to additional information of which they didn’t even know it existed. And while universities obviously exist to educate, there are additional benefits to them as well.  For one, it potentially saves their editorial team time; articles no longer need to be stuffed with categories and keywords. 

The higher accuracy of suggested content could also lead to a lighter load on supporting functions that are otherwise tasked with answering students’ questions. And then there’s the added benefit of keeping the visitors’ attention for longer. Something that’s perhaps not as important for universities as it is to other businesses, but it does help their SEM and SEO efforts.

For companies that make their money by keeping people on their websites for as long as possible, visiting as many pages in the process, is exceptionally valuable. Nowhere more so than in the media industry.

A Frosmo customer in this space had the additional problem that they weren’t just one news site, but a whole network that covered a vast spectrum of topics. Here, NLP was brought in to match and promote articles from one site to the other. Something that would’ve been exceptionally laborious otherwise.

The cross-promotion of articles across their network has the added benefit of introducing people to media sites they hadn’t visited before.

While both of these cases use NLP to automate the suggestions of articles with superior accuracy, Frosmo is looking to expand the use to other areas as well. If you have an interesting project that could benefit from this solution, let’s sit down and have a talk. As with so many things AI, the possibilities are endless.

Are you ready to start your personalization journey?

About the writer: 


Tom de Raad is the Product Marketing Manager at Frosmo. With 9 years of marketing experience for Finnish software startups, he’s business-educated and tech-curious. Tom joined as a Copywriter and is now using his skills to analyze and translate market needs into product features and vice versa. He’s currently exploring and mapping out the brave new world of personalization. He’s also winning the company’s Fantasy Football league.