Buying books online has never been as easy as it is today. But, with more than two million new titles published every year in the world, the discovery of new books is becoming one of the most fundamental problems in the publishing business.
How not to lose your way in the maze of bibliographies? How to make your way through the thousands of new items which every month flood bookshops on a rotation?
Apart from books themselves, all current methods of recommending books work online and all of these base themselves on client statistics and data; they suggest new items basing themselves on the principal that “clients who bought this item, also bought this other one”. For a publisher or a bookshop (digital or not), it is very difficult to obtain a sufficient number of users to create such statistics and the main dealers (Amazon, most of all) often do not share their data. Furthermore, they tend to push mostly for fast selling books, best sellers and fast sellers, often leaving the real treasure of a publisher, the catalogue, in the background.
A catalogue is a list of the long trail of books released by a particular publishing house from foundation to present day, comprising every success and every faliure since; it constitutes the identity of the publishing house, giving it shape.
Long sellers and classics emerge from this catalogue, books which withstand initail superficial reviews and manage to endure (at times regaining popularity after many years) the instatiable rotation of items in bookstores, displaying, for brief periods, thousands of new books which may never again sit on the shelves.
There is a conflict of interest, when whoever is selling you books is also responsible for helping you discover them; there are investments which need to bear fruit and marketing strategies which can be used. In theory, readers should always obtain whatever they want to read, not what the store wants to sell them.
TwoReads proposes a solution to this problem.
TwoReads is (and will be) a “discovery engine” for books.
It will show links between different texts, extracting bibliographical quotations for every book and showing influences and references which will help readers create their own reading path. It will analyse bibliographical references (i.e. the data which is buried in the book’s bibloigraphy, footnotes and book quotations) in order to recommend new books which may be of interest to the readers.
In this way, we can rediscover all the catalogues of the publishers, exploring the past (via the books which are quoted), connecting it with the future (via the books which they quote), giving readers a tool with which to navigate the spread of ideas and with which they can orientate their comprehension of the texts. In fact, as bibliometry and webmetrics show, the network of quotations between documents provides meaning.
In addition, it is posible to reccomend books basing oneself on books which the reader has already read.
In TwoReads reccomendations are neutral: they are obtained directly from the author of the book and readers can verify them themselves. There are no marketing strategies and the reader is no longer a mere figure in a statistic. The more he reads, the more these reccomendations will refine themselves and become personalised. We want to create a system which will offer serendipity, in order to discover books which are always interesting.
At the moment, we are in working on a pilot project with Il Saggiatore: we extracted, enriched and organised the links between 300 titles from its catalogue and developed a widget for online, catalogue-based reccomendation for further reading on each of its books. This is an example.
Furthermore, we are developing a platform where the user will be able to navigate via the books, finding new ones from these links. In TwoReads, it doesn’t matter if the book is physical object or digital one: the only relevant factor is that the book has been quoted or that it quotes other books. The system works principally with non-fictional books, such as essay collections, but it also includes high-quality narrative texts which are often quoted in other texts.
We believe in the existence of two readers – human and machine – and that the second can help the first in the search for a book, its discovery and understanding. Every book is a library, every book is a database.