Last time, I introduced some of the background of the research project that is the subject of this blog. I pointed out the diversity of approaches to data sharing and asking whether such systems can meet the various needs of users across multiple disciplines. This question forms the basis of our 2-year study of user experience in archaeology. We hope to arrive at an answer to this question by observing real, live users of digital content working in their “natural environment,” (a sort of Wild Kingdom episode about archaeologists with computers) with all the associated expectations, requirements, and limitations of their practice. My next few posts will present some of our initial findings. For those of you with the inclination to get into the nitty-gritty of it all, feel free to check out the project wiki, which documents our first kick-off workshop held at the end of January.
It is probably impossible to build a “perfect” resource. Every user has his own set of unique needs, working in his specific context. In order to meet the needs of everybody, a resource has to be extremely generalized. An example of this is today’s standard search engine, which works on the assumption that most search needs can be handled by some keyword searches of indexed content. By allowing them to do so in a straightforward way, and by not catering to any specific user needs, most people find what they want relatively quickly and therefore they are satisfied with the service. For researchers, this kind of generalization works to a certain extent, when people are looking for images or bits of information.
When it comes to creating research resources around primary data, however, the task is much more challenging. Who works with this kind of content? What are their needs? Even in archaeology alone, the diversity of needs is staggering.
For example, a 3-month study of search strings that linked to Open Context content resulted in the graph at the right. This classic “long tail” shows that most users are looking for something that could be classified as a person (names of colleagues, vanity searches), an object (images of objects for teaching, perhaps?) or a place (archaeological site names). These three categories account for almost half of the searches. The remaining 53% of searches were for all sorts of things that couldn’t be easily grouped into categories. While it is informative to us to know that people are seeking all sorts of information, we are left wondering—did any of them find what they were seeking..?
How digital resources are used and by whom is difficult to measure, and our study is attempting to find ways to better get at what motivates people to access primary content. Next time, I will lay out some of our specific observations on this topic, but first I want to describe the user group that’s informing this study.
We are studying this diversity in more depth by focusing on twelve specific producers/users of online content. These 12 individuals (the project’s “Community Liaisons”) came to Berkeley in January 2009 for an intensive workshop aimed at defining a baseline of the diversity of needs they collectively represent. The Community Liaisons form the crux of our study, as they each represent a complex mosaic of archaeological communities, experiences and approaches. They come from a variety of backgrounds, spanning field archaeologists to cultural resource managers to independent specialists. It became clear immediately that each of these 12 people has their own unique way of conducting their research and communicating their work to their respective communities. The diversity of tools that they use, alone, is staggering. Discussions revealed that some people are very interested in specific tools to share content within their community; others want a database tool to take to the field during excavation; and still others want broad project information linked to specific datasets, images and publications. Copyright, security and access concerns are pervasive, yet differ from person to person, depending on the context in which they work and their past experiences with sharing content.
How on Earth do you create a resource that serves all these needs?
I think the answer is: You don’t. There is no one resource that will satisfy everyone. Instead, you need to create ways for people to find and use content for their specific goals, working in their specific context. Rather than designing systems on the assumption that we know what’s best for each researcher, we’re trying to open things up to make it easier for customization of tools and content. This will give more options for different very specialized communities to create a context that works for their niche interests.
I’ll drill into this topic in more depth next time by asking: What is it, specifically, that archaeologists want to do with primary data and how do they want to interact with it?