Million Book Project
From Wikipedia, the free encyclopedia
The Million Book Project (or the Universal Library), led by Carnegie Mellon University School of Computer Science and University Libraries, aims to digitize a million books by 2007. Working with government and research partners in India and China, the project is scanning books in many languages, using OCR to enable full text searching, and providing free-to-read access to the books on the web. A pilot Thousand Book Project was performed to test the concept.
Twenty-two scanning centers are operating in India, including four mega-centers. Eighteen centers are running in China, including a mega-center in a free-trade zone to avoid customs delays with shipments of books from the U.S. Materials are also being scanned in Egypt, Hawaii, and Carnegie Mellon.
As of November 2005, over 600,000 books have been scanned: 170,000 in India, 420,000 in China, and 20,000 in Egypt. Roughly 135,000 of the books are in English; the others are in Indian, Chinese, Arabic, French, or other languages. Most of the books are in the public domain, but permission has been acquired to include over 60,000 copyrighted books (roughly 53,000 in English and 7,000 in Indian languages). The books will be mirrored at sites in India, China, Carnegie Mellon, the Internet Archive, and possibly other locations. The books that have been scanned to date are not yet all available online, and no single site has copies of all the books that are available online.
The million book project will provide a wide array of content, but one of its collection strengths will be agriculture. In partnership with the United Nations Food and Agriculture Organization, the United States National Agricultural Library, and university libraries with quality agriculture collections, the project is digitizing materials and developing plans for a knowledge network to improve rural community access to critical agricultural information.
Significant research is underway in the project, including OCR for Indian and Arabic languages and scripts. The research also includes developments in machine translation, automatic summarization, image processing, large-scale database management, user interface design, and strategies for acquiring copyright permission at an affordable cost. Indian partners have developed a translating and transliterating user interface. Partners in Egypt are developing an interface that supports annotation and highlighting. Partners in China have made remarkable progress on content-based image retrieval and machine analysis of calligraphic scripts. Carnegie Mellon has taken strides in machine translation and automatic summarization.
The National Science Foundation (NSF) awarded Carnegie Mellon $3.63M over four years for equipment and administrative travel for the Million Book Project. India is providing $25M annually to support language translation research projects. The Ministry of Education in China is providing $8.46M over three years. The Internet Archive has provided equipment, staff and money. The University of California Libraries at Merced funded the work to acquire copyright permission from U.S. publishers.
India, China and the U.S. agreed in November 2005 to join the Open Content Alliance (OCA), initiated by Brewster Kahle and the Internet Archive, because the goals of the OCA are consistent with those of the Million Book Project and the Universal Digital Library.
The bill of rights for the information society articulated by Jaime Carbonell, Director of the Language Technologies Institute at Carnegie Mellon, captures the vision of the Universal Digital Library: "Getting the right information to the right people, in the right timeframe, in the right language, with the right granularity."
Expectations are that the Million Book Project will accomplish its goal of digitizing one million books by 2007.
