A million books online for free  

By Ammar W. Mango

All library research rituals may become a thing of the past real soon. The process of running through drawers of indices searching for a subject has already been consigned to history, and in the near future the whole physical library concept may face the same demise.

The days of walking along thousands of bookshelves trying to find a book may become like a fairy tale for grandfathers to tell their grandchildren. Soon, libraries will be replaced by virtual ones; easily and freely accessible sources, from the convenience of your home, work, or school computer lab.

The Million Book Project, a name fit for such a legacy project, aims at digitising a million books in multiple languages, and providing them for free on the internet as part of what is called the "Universal Library."

The project, initiated by Carnegie Mellon University, is part of a bigger scheme to create a universal library that will capture all books in publication worldwide. This library is a virtual one that will exist on the internet for people to access freely around the clock.

The library will allow full text indexing using Optical Character Recognition (OCR) technology to preserve all written works in a digital format, instead of the more traditional and less reliable hard copy books that are prone to theft, fire, and decay.

At present, researchers in less developed countries, still dependent on regular libraries are suffering because of the limited number of resources. The universal library will end the problem and offer researchers access to a decent number of references, regardless of their geographic location. The online project will also provide education systems from countries around the world with the chance to make use of books that may not be available in their part of the world.

The project's scope includes six major components. One of these is the provision of a huge database for storage that will be mirrored at different sites worldwide. This is one of the major challenges facing the project. It is estimated that fifty petabytes of storage will be needed. This translates to fifty million gigabytes, which is beyond comprehension compared to most mega storage databases worldwide.

The bulk of the project involves the scanning of the million books into the virtual library using overhead type scanners. The data production will store images from scanners as Tagged Image File Format (TIFF) at a resolution that will provide for reading the book directly off the screen.

To allow for full text search, which is an important feature of the project, OCR will be used. However, the OCR version of the text will not be accessible to users and will only serve search purposes.

Since the library is going to feature books without violating copyright laws, it is expected to include relatively older books, in addition to newer ones that grant permission to the library for book inclusion.

The project which started on August 2001 is well already underway, with over 600,000 books already scanned. Scanning has been taking place mainly in India and China, with some books scanned in nearby Egypt. While most references are in the English language, some are in other languages, including Arabic.

The website for the library is already up and some of the books are already available on the site for searching by users. The site address is http://tera-3.ul.cs.cmu.edu/. There are already four books in the archive about Jordan, in the English language. In the Arabic language, there are only two references that I have managed to find. Visitors can search by title and author, as well as perform a full text search. Users can choose the language of the search and the subject sought.

Funding and support for the Million Book Project are coming from many agencies, companies, and institutions from around the world. The American National Science Foundation has provided the necessary funding for the equipment used. Manpower resources are mainly provided by India and China to cover scanning, indexing and hosting.

There are also several existing sites that provide books in an electronic format via internet. These include www.questia.com, children's books on-line www.childrensbooksonline.org, international children's digital library www.icdlbooks.org, project Gutenberg www.gutenberg.org, and the internet archive www.archive.org. Access to all of them is open and free.