07.05.2015

Tekstum: Big Data can also help us choose books

Image | Daniel Y. Go

Nowadays, choosing a book is no easy task. The socioeconomic crisis has spawned a very competitive market in which consumers do not have as much purchasing power as they once did and are not able to take home as many books as they would like to.  Bookstores have transformed into showcases where the most important thing is to have the most eye-catching presentation or the biggest stacks of volumes, in this way luring everyone who walks through the door.

When shopping, the indecisive reader has several ways to choose a book. He is most often influenced by recommendations from his close circle of family members and friends, or even the bookseller. Another option is to look at sales volumes, such as those posted during World Book Day. It is also common for readers to demonstrate interest in all those books that have received awards. Finally, they may try to find out what books are best for them based on the information available on the Internet, including specialised reviews, opinion forums, and the social networks.

TK2

Image | Tekstum

Now readers have a much more powerful tool available when it comes to choosing a book: Tekstum. This is a platform that, using an algorithm based on Big Data and Natural Language Processing (NLP), analyses all the existing information on the Internet on published works.  In order to do this it employs Sentiment Analysis, a new branch of study that, by identifying key words and specific expressions, seeks to determine the polarity of any bit of content published on the Internet. The other parameters allowing Tekstum to produce a global evaluation of each work are the book’s “social engagement” and its author’s social reputation.

Tekstum represents a tool boasting great potential, allowing publishers to gauge the public’s tastes and readers’ profiles in greater depth, an added value that comes in addition to the information generated by sales figures. In this way Tekstum can serve to reduce uncertainty about the public’s reaction to a book, centre the business model on demand, and discover new national and international values.

TK3

Image | Tekstum

Since 23 April the platform has featured a beta version presenting a selection of the books with the best ratings based on the Tekstum algorithm. During this first stage the tool will recommend similar books for each work analysed, and will offer all those who register a free analysis of whatever book they choose.

Tekstum is an initiative backed by Marc Santandreu and Marc Martínez, economists with experience in the publishing world, and forms part of Incubio, the incubator specialised in Big Data.

What about you? Do you already know what book you’ll buy next?