Acknowledgements

Gregory Colati
Director, Digital Collections and Archives, Tufts University

The project began, as so many of these things do, at lunch. Back in 2001, Nancy Richard, the Director of the Bostonian Society Library and I were having lunch and discussing the recent completion of another Tufts IMLS project that used maps to locate both words and pictures about the history and topography of London. Nancy wondered aloud if it would be possible to do something similar for Boston, after all, she had one of the most extensive collections of early urban photography in the United States, and Tufts had the infrastructure and know-how to create an exciting way to deliver it electronically.

The key issue as we saw it, was to find a way to combine the three most used resources in the library (photographs, maps, and city directories) in a way that would transcend their use in the physical world.

At that time the Bostonian Society did not have either a internal or public electronic catalog of its holdings. So we began with a modest project to begin cataloging the collection into a collections management database. Nancy was able to secure funding from a private donor to purchase collections management software and to start cataloging her collection. At the same time, Tufts was very fortunate to be selected for a National Leadership Grant from the Institute of Museum and Library services to try to build a system that could create a context for the people and places of the city in time and space without having to laboriously geo-reference thousands of photographs and millions of directory entries.

We began with photographic prints, bound paper city directories and large format maps and atlases and an idea that somehow we could combine them in ways that would support use and users. The original grant narrative (pdf) said that we would seek to "demonstrate that it is possible to organize data by spatial location and then link this data to many other pieces of data that share the same location without having to extensively catalog them by subject, name, or other subjective terminology." While this was true in one sense, what we ultimately discovered was much more valuble. That is, that we could link all sorts of data to a subjective place name, as long as we could anchor that subjective place name at some point to a particular location.

Converting these resources into digital content that would serve the needs of the project and, ultimately, those who use the tools we provide, took numerous steps and involved people working across the city, the country and the world.

At first glance, the project personnel list seems to indicate a massive production system with armies of people working on the project. In truth, very few people were actively involved for the entire two years, and many would "check-in" and out was they were needed or available. Keeping it all running smoothly was Project Coordinator Jessica Branco. Without her organizational skills, creative vision,and technical understanding, the project would not have succeeded.

In all libraries, public services ride on a technological backbone. For this project, that backbone was provided by the department of Academic Technology (AT). The development of the Boston Streets project ran in parallel with the development of the Tufts Digital Library and repository. AT programmers and DCA technology staff worked together with the content specialists to ensure that the tagging and cataloging decisions made were ones that would support the things we wanted to do.

The real hero of the project was Peter Wilkerson of Asheville, NC. Peter was brought on to the project in the second year as we struggled with, and finally solved the vexing problem of how to automate the linking of cataloged images, addresses listed in the directories, and geospatial points, and then how to display this information at the appropriate locations on the historical and modern maps of Boston included in the project. This was the key relationship that we hoped to demonstrate in the
project and we are very excited about the potential of the results. Peter and Robert Dockins (a Computer Science graduate student who also joined us in the second year) were instrumental in developing the solution. By regularizing and manipulating the data using regular expressions and PERL scripts, instead of hand encoding, harvesting, and plotting each piece of information on the maps by hand, workflow become efficient, accurate, and highly automated.

With the ability to join together, filter and deliver sizeable amounts of information on demand, we are able to show that tying data to geographic points can lead to increased
knowledge and the ability to discover information about something even if you know nothing else but the spot on the earth. Building layers of context around a point, a picture,
or a person benefits the cataloger, the researcher, and the casual user.

The photograph catalogers at the Bostonian Society, Carrie Allen and Anne Vosikas, developed a cataloging workflow and manual that became not only the standard for the project, but was adopted here at the DCA as well as a model for cataloging images.

Fancy automated manipulation was not the only route to good data, the initial directory scanning by our student staff, using a Minolta PS 7000 planetary scanner provided the true raw material for the project. Page images were sent to Digital Divide Data for double keyboard data entry before being returned to us for tagging and regularization. More information about these processes can be found in the technical documentation section.

Finally, I would like to once more thank everyone who participated in the Boston Streets project. It is a wonderful resource and another step in the evolution of reference and research tools for libraries and archives.

Gregory Colati
Medford, MA
July 2004