MW-photo
April 15-18, 2009
Indianapolis, Indiana, USA

Putting Museum Collections on the Map: Application of Geographic Information Systems

Megan Heckert, Avencia Incorporated, USA

            http://phillyhistory.org, http://www.muralfarm.org          

Abstract

On-line mapping tools are quickly becoming ubiquitous, and have much to offer museums for both collection management and outreach. This paper explores potential applications of Geographic Information Systems (GIS), focusing on how two organizations, the Philadelphia City Archive and Philadelphia’s Mural Arts Program, have used GIS to expand their on-line presence and reach wider audiences with collections of historic photos and information on public art.

This paper focuses on how the geographic approach has spurred the growth of these sites into valuable assets for collections management and beloved sites for users. It explores how geographic technologies can be applied to both collections management and outreach for archives and museums to expand their reach and provide patrons with a new lens through which to view exhibits and collections.

Keywords: mapping, GIS

Introduction

Place matters. When poet-farmer Wendell Berry said, “If you don’t know where you are, you don’t know who you are,” he spoke to the deep importance of place in making meanings. Where an event happened, an object was created, or a photo was taken can have significant impact on how that item, object, or photo is understood.

Philadelphia City Archives

Philadelphia’s City Archives, managed by the Department of Records (DOR), contains an estimated 2 million photographs, some dating to the 1860s. Photographers assigned to City departments such as the Departments of Streets, City Transit, and Public Works originally took these photographs in order to document public works projects. 

Facing several challenges, including a rapidly aging collection, limited public access, and an ever-tightening budget, DOR partnered with Avencia Incorporated, a local software company, to develop PhillyHistory.org. PhillyHistory.org was established in 2005 as a means of preserving and organizing the Philadelphia City Archives’ collection of historic photographs. The project encompasses a mass digitization and preservation effort as well as the development of the PhillyHistory.org Web site (www.phillyhistory.org). PhillyHistory.org serves as a Web-based digital asset management system, enabling both management of metadata about the images and public search of the collections based on location, address, time period, and keyword. An interactive map shows the locations of all available photos near an address, intersection, or neighborhood, helping users to identify photos and understand historic settings in their current contexts.

Mural Arts Program

The walls of Philadelphia are covered in murals – more than 2,800 of them. Established in 1984, Philadelphia’s Mural Arts Program’s (MAP) mission is to engage in art education and community public arts collaborations, and to increase public access to art. The reach of these services, however, is limited to local residents, and the opportunities are restricted by organizational capacity. To overcome these limitations, MAP developed an on-line catalogue of its murals, organized by location. Muralfarm.org enables users to find murals based on their location in the city and to access photographs and information about them. MAP’s murals are designed using a collaborative, community-driven process, such that each mural is the product not just of the artist or artists who drew the final design, but also of the community whose wall (or walls) it graces. GIS offered MAP a way to display and share their murals in their community context, painting a richer picture than might otherwise have been possible.

Evolution of the Technology

PhillyHistory.org and MuralFarm.org both rely on the Sajara software framework developed by Avencia Incorporated, which has evolved over time to meet the needs and expand the functionality available to the projects. PhillyHistory.org was the first to be developed and originally consisted of collection management features, simple public search options, and an e-commerce system. Public users of the system could search the collection based on an address, a neighborhood, keywords, and dates. A small portion of the images was unable to be georeferenced, and an additional search option included images that did not have a location.

Very early in the project, several challenges presented themselves. One of the most significant was developing the system to geocode the images (i.e., assigning a geographic location on a map to each photo). Most images are not indexed to a specific address, but rather have a description along the lines of “east of X Street looking towards Y”. This was addressed through the development of a parser that would read through the descriptions, picking out addresses based on common configurations. Geocoding was further complicated through changes in street names over time. Street name changes are accommodated through the use of street aliasing based on information maintained by the Philadelphia Department of Streets. However, we knew that the resulting system would not be perfect, so the project team implemented a feature that enables site users to provide corrections when they have knowledge of an image that has been geocoded to the wrong location.  PhillyHistory.org has developed a small but dedicated subset of users who regularly provide location corrections for photos. These corrections are often connected to rivers, bridges, and railroad lines, locations that can be notoriously difficult to geocode but that many users often have personal knowledge of and  can assist with accurate geocoding.

As digitization of the Archives photograph collections continued, questions began to arise about the potentially controversial nature of some images. With the Web site available to the public, images from the medical collection, for example, were deemed inappropriate for display. As the project was also about preservation of the negatives, though, it was important that these photos be included on the management side, and the ability was added to turn the public view of an image on or off to accommodate this. Other images, though potentially troubling, such as a 1928 theater marquee depicting Al Jolson in blackface were seen as being important historic records. To address potential concerns over these images, a blog was created to provide additional context for these images (though blog entries are not limited to controversial topics).

From its conception, PhillyHistory.org was also seen as a potential source of revenue for the Department of Records, and an e-commerce component was built to enable members of the public to purchase prints from the collection. This has proven quite successful, and even at the relatively low cost of $10 and $20 per 5x7 or 8x10 print, the project has generated $36,870 in revenue, with additional income from the sale of high-resolution digital images and large framed prints from an art exhibit. Purchase options are limited due to orders being fulfilled in-house, and we feel there is potential to increase revenue considerably beyond this by offering additional services. A project is currently underway to enable third-party order fulfillment by a vendor that would provide many additional product offerings.

Adaptations and updates

Over time, the system has continued to change. The search user interface has been updated several times to incorporate feedback from users and new interaction technologies such as AJAX. One example is the location of the address search dialogue. In the earliest versions of PhillyHistory.org, the address and neighborhood searches were both coupled with the map interface and only visible when the appropriate menu was clicked on. It seemed an appropriate choice to keep the location parameters together. But watching public users navigate the search page quickly revealed that this had not been the right decision. Many people were typing addresses to begin their searches, but they were typing them into the keyword dialogue, which was at the top of the page. Based on this observation, the search page was reconfigured with the address dialogue always visible at the top of the page.

Additional updates have revolved around users’ ability to interact with the collections. Users can save images to a ‘favorites’ collection and assign custom tags, e-mail an image to a friend, save searches, and subscribe to search RSS feeds for updates as new images are added to the system. As Web-based geographic technologies have continued to improve and become more familiar to the public, additional features have been added to the site, including the ability to view search results (or the entire collection) in Google Earth and to view RSS feeds in GoogleMaps using GeoRSS.

The Mural Arts Program began work on Muralfarm.org in 2007, introducing a new set of challenges. Chief among these was that the nature of the assets was quite different. With PhillyHistory.org, each photo represented an individual asset. For the Mural Arts Program, the assets are murals, each of which might have several associated images. Because the system was expected to work for MAP for database management as well as public outreach, it was important that each photo have its own metadata separate from the metadata of the mural itself. The database and asset management interface had to be redesigned to accommodate this difference.

Not all attempts to incorporate new features have been equally successful. Mural Arts also experimented with Google Streetview as a means of enabling users to view existing murals in their larger context. While it was an interesting idea and compelling feature in theory, the reality left quite a bit to be desired. The difficulty was that Streetview would be configured based on the address to which the mural had been geocoded. Unfortunately, we were not able to configure the system to ensure that the Streetview popup would actually point directly to the murals, which we found were often around the corner from the view provided, and in some instances no amount of navigation would bring them into view. Mural Arts toyed with the idea of turning off the feature altogether, but in the end settled on adding a disclaimer indicating that users may have to move around a bit to see the mural. (Note: Another take on using Google Streetview to add context to photos was described by Paul Hagon (2008) who built a mashup using geotagged images from the Powerhouse Museum available via Flickr Commons.)

The Impact of Geography

For both the Department of Records and the Mural Arts Program, the ability to manage assets based on geography has been a valuable aspect of the projects, due in part to the nature of the assets. The DOR’s photos are valuable in providing a historic record of a specific place. For the Mural Arts Program, location is just as much a part of the mural as the actual image. For site users, the ability to search assets by location is central to their experience of the Web sites.

Since all searches have a geographic component based on map extent, it can be difficult to quantify the instances where geography was intentionally incorporated versus when the default city-wide map was used. However, nearly 40% of searches in PhillyHistory are performed based either on an address or on a neighborhood, clearly demonstrating that users are relying on location as a significant component of their use of the site. Muralfarm.org is scheduled for public release at the end of January 2009, as this article is being written, so similar statistics are not available at this time.

While we see GIS as having a very important role to play for these two applications, it may not be the most appropriate approach for every collection. Location might not be as vital a component for collections such as portraits or media that reference fictional lands. By contrast, the incorporation of location may be particularly valuable for cultural heritage collections. Research by the Dutch government indicated that the public’s main interest in on-line cultural heritage is family and local history (Liberge & Gerlings 2008), and our experience with PhillyHistory.org has certainly supported this.

Conclusion

When PhillyHistory rolled out, we were surprised by the comments we started getting from site visitors. As public use of the system has increased, we have been amazed by both the range of users and the depth of personal reactions users are expressing through the site. In addition to historians, genealogists, librarians and researchers, the site has been used by apartment rental companies looking for historic images of their properties to decorate their new buildings, authors researching settings for their historical novels and screenplays, as well as curious members of the general public. Visitors can submit comments through the site, and many describe how they started their visit by searching the neighborhoods where they or their parents or grandparents grew up. They are sharing very personal reactions to the images, illustrating the stories of places that are very dear to them and that anchor them to their city and community’s history.  “You helped me remember my past,” wrote one site visitor. “I see parts of the city as I was growing up and think of how things have changed. Thank you.” “It feels like I found jewels,” wrote another. A visitor wrote, “It was so great to see how things were and looked so many years ago in a city that many people who grew up here cannot let go of after so many years” – illustrating the deep connection people feel to the City of Philadelphia and the places depicted in the photos. And while we haven’t actually put him to the test, we were quite thrilled when the former editor of the Philadelphia City Paper actually wrote, “This is the best use of taxpayer money I’ve heard of in a long time. I’d even be willing to pay more taxes, if it would speed up the image uploads.”

References

Hagon, Paul. 2008. Powerhouse Street View Mashup. http://www.paulhagon.com/blog/2008/08/19/powerhouse-street-view-mashup/. Accessed 11/20/2008.

Liberge, Leila and Job Gerlings (2008). Cultural heritage on the (geographical) map. In J. Trant and D. Bearman (eds.) Museums and the Web 2008: Proceedings (CD-ROM). Toronto: Archives & Museums Informatics. Published March 31, 2008. http://www.archimuse.come/mw2008/papers/liberge/liberge.html.

Cite as:

Heckert, M., Putting Museum Collections on the Map: Application of Geographic Information Systems. In J. Trant and D. Bearman (eds). Museums and the Web 2009: Proceedings. Toronto: Archives & Museum Informatics. Published March 31, 2009. Consulted http://www.archimuse.com/mw2009/papers/heckert/heckert.html