A&MI home

158 Lee Avenue
Toronto Ontario
M4E 2P3 Canada

ph: +1 416-691-2516
fx: +1 416-352-6025

info @ archimuse.com

Join our Mailing List.



published: March 2004
analytic scripts updated:  October 28, 2010

Creative Commons Attribution-Noncommercial-No Derivative Works 3.0  License


Museums and the Web 2003 Papers

Why Not Google: Is There a Future for Content Aggregators or Distributed Searching?

Willy Lee, Minneapolis Institute of Arts, USA


Considering the power of modern search engines, are the goals of content aggregation and distributed searching still relevant? While there is a strong case for the usefulness of such tools, can museums expect users to take the time to learn how to use these tools while sites like Google may be good enough? What do museums need to do to set themselves apart, or what can be done to make the Google experience better?

Keywords: content aggregators, distributed Search, context, Google

Since the late 1990s, the art museum community has talked about the potential for the Web. Protocols like z39.50 could distribute searching, allowing a single query to find information from a large number of databases. Multi-museum collaborative WWeb sites could bring greater understanding to an artist's work. With a medium as cheap as the internet, we could throw open the file cabinets and set information free.

So where are we now? z39.50 and other federated search technology leans heavily towards library science and seems unfriendly to the average user. Two types of other museum collaborations have cropped up: small scale and subscription based.

Small-scale collection-based collaborations lack the breadth to be useful to a global community. While sites like ArtsConnectEd (http://www.artsconnected.org) serve geographic regions well, such limited scope makes them less useful to general research. Projects like CHIN's virtual museum (http://www.virtualmuseum.ca) are useful, but only if you are looking for a work in a Canadian collection.

Projects like AMICO (http://www.amico.org) and ARTstor (http://www.artstor.org) create this breadth by providing a simple way to search deep within disparate collections. These collaborations can also help secure rights from artists who may still hold copyright over their images. These sites by their nature are restrictive: restrictive for users since they carry high subscription costs, and restrictive for content providers because of their high technological barriers to entry. Only the largest and best funded museums and educational institutions can afford to jump through the technological hoops required to get their collections exported into such systems.

News flash: federated search has been solved. Modern search engines have really taken the concept of federated search and reshaped the essential question. Instead of "How are we going to search all this stuff?" the question becomes, "how can we get the good stuff to the top?"

Advanced features in google allow you to limit a search to a certain domain or subdomain. When "site:org" is added to a search, google limits the returns to pages that are hosted in .org domains. If museums truly adopt the .museum domain, a search like Rembrandt site:museum would only return content from museum Web sites. This can be further refined to Rembrandt site:art.museum to limit returns to pages from art museums. In a limited sense, this works today, though not many pages have been indexed under .museum names as few museums have adopted them as their primary domain names.

It seems pretty unreasonable to expect the museum community to market this idea to the masses. "Hey everyone, if you want to limit your search to just art museums that use the .museum domain type this in…" A more workable solution would be to have museums add search boxes to their sites to add this to a google search automatically. This could go even further if these boxes were branded in some way so that after seeing one on the Metropolitan Museum of Art's site, the user may recognize the same form on another museum site and properly identify it with the function it serves.

This approach also lowers the barrier to entry for museums to participate. Many of the world's cultural heritage institutions run on slim budgets and can barely scrape enough together to put up a Web site. The requirements to enter are lowered to simply having a .museum domain name and being able to post a Web page. No fancy databases and XML exports are required.

Three of the more popular art history sites are not affiliated with museums or educational institutions at all. Artcyclopedia (www.artcyclopedia.com), artchive (www.artchive.com), and the webmuseum (www.ibiblio.org/wm) are all efforts by individuals to provide access to works of art from throughout the world. These sites are appearing higher in google returns than museums are when searching for an artist by name. These resources are available  free and from anywhere. They do well because they have a broad scope and fairly lengthy biographies and essays. Where is this content coming from? Much of it is copy stand photography and scans from art history books. The text content appears to be rekeyed from other published materials. Though some may see this as questionably legal in regards to copyright, the fact of the matter is that these sites continue to prosper.

These are the sites that get used in k-12 schools when students are writing papers. These are the sites students are looking to when they are looking for art images. Why? Simply because there are few barriers to access. There is no subscription cost, there is no need to go to the library. On top of that, we have teachers praising these sites without regard to the authority of the information. Beyond this, these sites offer two things that museums often shy away from: breadth outside a single collection, and context in an ouvre - content and context that is being liberally quoted from Art for Dummies and similar sources and presented by these individual

Of the top 200 search terms at http://www.artsmia.org, 27% are artists' names, while 16% are cultural groups or geographic terms. Information about these topics is the type of information that most collection databases lack. Certainly, the names and countries show up on those records, but as far as the search engines are concerned, is not terribly relevant. It doesn't even seem all that relevant to the user as well. In isolation, these records do little to further the understanding of art.

If museums really want to bring content to the average internet user, there must be greater emphasis on the contextual information and less on the database. Many of the popular sections of http://www.artsmia.org use a thematic structure to paint the context of a theme and flesh that out by allowing the user to explore a large number of database materials. Modernism (http://www.artsmia.org/modernism/) and Unified Vision (http://www.artsmia.org/unified-vision/) appear at the top of google returns for modernism and prairie school because of this. These sites provide the contextual wrappers  these collections, but then allow database access to open the work to individual exploration. These are the sections that draw the most traffic, and these are the sections that draw the most praise.

As more and more people turn to the Web as their primary source of information, museums need to direct their limited budgets towards projects that reach greater audiences. By leaving federated search-to-search engines like google, museums can focus on delivering the context that the public seeks. Without context to lead into collections, the pieces stay isolated and underutilized.