At BooksAI, we take a more holistic and neuroscience-inspired approach to teaching machines how to feel books as humans do that is not based on traditional techniques.
In the future, it could be the way that bookstores find new books for you, the way librarians make connections between books that Amazon’s and Google’s algorithms don’t see, and the way that enterprising readers find something new and unexpected.
Beyond "Also Bought"
In the future, it could be the way that bookstores find new books for you, the way librarians make connections between books that Amazon’s and Google’s algorithms don’t see, and the way that enterprising readers find something new and unexpected.
Beyond "Also Bought"
All of these methods have the effect of the same popular, top-selling books being recommended repeatedly.
Traditional recommendation systems are still based on sales rank, purchasing history, other people's opinions even though about a third of online reviews are fake.
It works for some, but not for others…
How about your personal taste?
Every author has own style, attitude, mood and tone. When you read several books by the same author, you become accustomed to the author’s style of writing and sometimes you look for authors with a similar style.
It is rather often that a tone, mood, intonation and a lots of other hidden elements of style happen to be much more important than the genre or even the plot.
It's not enough for a modern recommendation system to just answer the question "what it is written about", it has to understand "how it is written” and this is a very tricky thing.
Because it is not about search, keywords or publishers metadata.
It is NOT about words at all. Surprisingly...
One million new English titles are being published every year and readers can choose from a back-catalog over 30 million titles.
There's no problem finding the books; it's finding good book that's the problem..
We always asked our friends, bookstore assistant or librarian
"Who writes like" or "What book is like"
But there is no one person in the world who could read 30 000 000 books and make recommendation.
Furthermore, most of these books remain unread. Not because they are bad, but because lack of marketing or “lack of luck”.
Since most of these books have never been read; therefore, they cannot be recommended.
Simplicity
There isn't too much of time for people to spend on traditional book discovery anymore.
The discovery processes must be simplified dramatically.
We should ask "Siri-or-something" one simple question.
- what book is like Eat, Pray, Love? I liked it very much and I am hungry for more.
- who writes like Somerset Maugham? I like his style and attitude.
And we will get the right answer
Serendipity
One of the joys of traditional bookstores was serendipity.
At some level we'll need to create our own serendipity in the e-book era.
We'll probably need to do it more consciously, by going outside our zones of comfort and the recommendations of people we trust.
We call it "The Zen Of Book Discovery" -because it is all about
relevancy, simplicity and serendipity at the same time
....paradoxical and slightly mad.
But at first, we have to teach machines how to “FEEL” books as we do.
How we do it? Check out our thoughts on technology and metadata.
Want to read more? Contacts us for additional information or demo.