Read all about it! Virtual library assistant helps readers find books based on their moods and preferences
A Polish tech start-up has created a tool for recommending books using artificial intelligence.
Acting like a virtual librarian, its software points readers toward books that they might enjoy, based on their individual mood and preferences.
The start-up was co-founded in 2016 by Małgorzata Sieniewicz (R) Urszula Witkowska (L), both of whom have a background in book sales and reading more generally.
With thousands of books published every year, choosing one to take on holiday or to give to a friend can be tricky. Even with reviews in magazines or online, the choice can be overwhelming.
Polish startup Literacka (which means “literary”) wants to make things easier for readers. Based in the city of Olsztyn, in north-eastern Poland, it was co-founded in 2016 by Małgorzata Sieniewicz, the company’s CEO, and Urszula Witkowska, the director for development, both of whom have a background in book sales and reading more generally. Its team of ten includes IT architects, researchers and a project coordinator for the UK.
On its website bestbookadvisor.com, users are asked a series of simple questions about what kind of book they are looking for -- something funny, romantic, gripping, and so on.
Its project entitled “Automatic Reviewer – Intelligent Book Recommendation System” uses deep learning technology and natural language processing. These enable its software to mimic the workings of the human brain and read, understand and derive meaning from human languages – in this case, the content of a book.
Literacka’s tool finds similar plot and narration characteristics, which it uses to create its recommendation system. Based on the reader’s preferences – some people want a romantic fantasy novels set in the past, while others might feel like a modern thriller with a sense of humour – it guides them to books that they might find interesting.
With an overwhelming choice of literature on offer today, the software acts like a virtual librarian pointing readers toward books that they might enjoy.
On its website bestbookadvisor.com, users are asked a series of simple questions about what kind of book they are looking for -- something funny, romantic, gripping, and so on. They can also choose whether they are looking for a historical or modern book. For each of the eight criteria, readers can pick one of three levels, which roughly correspond to “not”, “somewhat” and “very”.
Having started with Polish books, the company is currently expanding its database of books published in English. This will enable the website to recommend books in English, too.
Fiona, who will be judging the BLEF literary competition at the Poznań book fair next month, can also read books with dizzying speed and using algorithms can catch the advanced features of a story.
The company’s research and development project is co-financed by Poland’s National Centre for Research and Development, with support from impact investment venture capital fund Simpact.
