Since planning began in 1985, the Perseus Digital Library Project has explored what happens when libraries move online. Two decades later, as new forms of publication emerge and millions of books become digital, this question is more pressing than ever. Perseus is a practical experiment in which we explore possibilities and challenges of digital collections in a networked world.

Our flagship collection, under development since 1987, covers the history, literature and culture of the Greco-Roman world. We are applying what we have learned from Classics to other subjects within the humanities and beyond. We have studied many problems over the past two decades, but our current research centers on personalization: organizing what you see to meet your needs.

We collect texts, images, datasets and other primary materials. We assemble and carefully structure encyclopedias, maps, grammars, dictionaries and other reference works. At present, 1.1 million manually created and 30 million automatically generated links connect the 100 million words and 75,000 images in the core Perseus collections. 850,000 reference articles provide background on 450,000 people, places, organizations, dictionary definitions, grammatical functions and other topics.

Our audience is substantial. In April 2005, we served more than 11 million pages to more than 400,000 unique users. These users come from around the world and from many backgrounds. Students, teachers and researchers use Perseus, but more than half of our audience comes from the general public.

We invested from the start in standards compliance and best practices. As a result, all objects of persistent value that we have collected can now find a permanent home in Tufts' institutional repository, where they will remain useful for years, decades and even generations to come.