Intellogo Brings Machine Learning to Book Recommendations

While the technology industry is abuzz with new opportunities created by advances in artificial intelligence -- from intelligent web search to voice recognition -- book publishing has yet to feel the full impact of AI. Neil Balthaser intends to change that through his machine learning software Intellogo. Intellogo uses technology that can analyze and understand the contents of a book in order to provide better book recommendations to readers. The technology can also identify reader behavior trends that can inform book publishers’ content creation strategy. In the following interview, Balthaser explains how Intellogo’s software can help publishers streamline book discovery and how AI will transform the reading experience into a conversation between the reader and the book.

Who are you?

Here at Intellogo, we do Big Data analysis on the content in books and reader behavior. We use machine learning, (a subset of artificial intelligence) to crack open books and understand their content: themes, writing styles, pacing, emotions, etc. We then use that data to better match books with readers.

Intellogo is equipped to understand what is actually in these books. Our system is trained to understand content in a more conceptual way, without relying on keywords being present. It thinks in more human ways – we've developed clever AI that interacts in a natural conversational manner, closer to the way we normally do.

What problem are you solving?

Readers are overloaded with content and it's constantly growing. Our reading habits haven't really changed -- we are still limited by how much we can read in a day.

We need to understand the reader and content better. We also need to create new opportunities to bring in new readers. Better technology integration solves a number of challenges. Similar to the Netflix model, Intellogo's analytics recommends content, and identifies trends so publishers can put the right focus on what content to publish next.

How did you come up with this idea?

I’ve been working in disruptive rich media technologies for over 20 years, from creating early developments in Macromedia's Flash technologies to working alongside Future Publishing’s creative team to craft one of the first interactive digital magazines for the iPad called MacLife. These experiences led to an opportunity at Barnes & Noble to create their digital publishing platform.

In 2013, I returned to the role of entrepreneur, founding Intellogo Inc. with the goal of connecting people’s curiosity and passion with humanity’s greatest gift, our written history.

What are the most important trends in publishing today?

We must understand and identify what is the primary business problem that all other problems are subordinate to. In the case of publishing it should be everything that is market data driven: connecting the right content to the right person. Focus on resolving business needs, instead of being product driven, and then hope there is a market for your solution.

Who is your competition?

There are very few competitors in our space. Machine learning has been more readily adopted in other markets than it has in trade publishing. There are a few players such as Trajectory and Amazon's X-ray that leverage technologies but don't focus on the same aspects of book content as we do. Retailers such as Amazon and Kobo are known for using Big Data to help drive sales. Booklamp, acquired by Apple in 2014 was the closest competitor to what we are doing.

What's next?

The next opportunities we see use conversation as a platform: using AI-driven interactive conversation, in natural ways we normally talk, to help with book recommendations. You'll be seeing Intellogo used as a Bot, either on messaging platforms like Slack, or through book vendors helping you discover the right books you were hoping for.