Model Comparison Paper
My basic idea is to contrast the works of Dervin and Belkin, and making the claim that both have the same idea, but take different stances as to what the most important aspect of HIB is. Dervin’s core idea is human understanding, Belkin’s is “help us (the system) help you”.
- Introduction
- Dervin’s model
- Focus on the user’s mental state
- Is a justification for a very good information rich interview technique
- Does not account for the system at all. Once data is gathered that is it
- Belkin’s ASK
- Based on the same precepts as Dervin’s model re. the user- vocabulary and some aspects are different, tho
- Also contains a view of the system and the generator and how the user interacts with a system
- Combined with Dervin, emphasizes that a system’s understanding of a user is paramount in providing the correct result to a user
- Dervin’s model
Dervin v. Belkin- Two Different takes on “User-Centered Design”
Introduction
I think it is fair to say, at least from the survey of the field explored in this current context, that most of the developments in the Information Sciences in the past few decades have been based around the idea of “User-Centered Design”- the idea that regardless of how well a system is engineered from a data-retrieval standpoint, if it does not match the needs of the user, it’s not going to be very useful. We’ve seen a lot of theories about how to tie this idea in to the development of systems, but I think two of the most powerful and useful models are Dervin’s sense-making model, and Belkin’s ASK model. Part of their power, in my opinion, comes from the fact that while they are both similar in many ways, they approach the user-centered problem from different sides- Dervin’s approach focuses deeply upon drawing out the internal thoughts of the user to construct a clearer picture of how the user operates, while Belkin focuses more upon the systems applications of such understanding- once we know what the user wants and how they work, how must a system take this knowledge into account? By using these two frameworks in concert, one can develop a powerful understanding of the interaction between user and system.
Dervin’s Sense-Making Model
Dervin’s method for understanding human sense-making is mind-numbingly simple- the appropriate way to look at how people make sense of the world around them is by conceptualizing any event requiring a decision as a “gap.” The decision of what shoes to wear? That’s a gap. Grape-nuts or corn flakes? That’s also a gap. Whether or not to even get out of bed? Gap. The idea is that at each one of these junctures, you are faced with a decision, and must construct some manner of bridge to get between the question and the answer.
At first, this method may seem horribly vague, and the truth of the matter is that it is- this is by design. By making this description so vague, we can fit a large number of very disparate activities inside of this framework- which ultimately is the intent. A model would not be very useful if it had to be radically restructured for each event it was applied to- it wouldn’t be a model of anything more than a very specific system. The other advantage to this simplicity is that it emphasizes the point that people do not solve problems based on their ethnicity, or any number of other statistics that have historically been used to describe a population (and to some degree still are in poorly-disguised marketing ploys).
Where this actually does become very useful, though, is in the idea of the micro-moment interview. This idea, which comes from Dervin’s model, is an interview methodology intended to take stock of how the user perceives the problem directly in front of them.
This model, though a good one in some situations, is severely limited in others. This model focuses entirely on the user’s cognitive state at a moment in time- nothing exists external to this. There’s no place in this model at all for the system.
Belkin’s Anomalous States of Knowledge
Once we have this picture of how the users interact with a system, though, what do we do with it? Belkin’s ASK model provides a better picture of this situation.
It is worth note that Belkin’s model does indeed have a wider scope than Dervin’s model- where Dervin intends only to focus on the “sense-making moment” as the basis of her theories on HIB, Belkin gives much more focus to the system that this information is stored in. In this sense, the model is slightly more rigid from the user perspective, and as such is not as useful when thinking about how to get a sense of the user’s need and how humans naturally go about solving their problems, but it does provide a good sense of what to do with this information once it has been acquired.
Belkin introduces the idea of an ASK, or Anomalous State of Knowledge, and posits that this the center of a user’s information problem. This concept serves the same primitive function in Belkin’s model as the gap does in Dervin’s (it represents what must be resolved before we can claim that an information need has been satisfied). Furthermore, it places the same thing at the center of understanding HIB- the cognitive state of the user. However, by conceptualizing this problem (or “anomaly”) as a problematic user state rather than a gap, it allows us to better conceptualize the system’s imperative task of resolving the user’s problem, and frames the whole thing in a constructivist theory of information retrieval. When a user poses a query to a system, what the system returns in response to the user’s query (whatever form it may actually take) is taken to be some kind of text, which transforms the user’s state of knowledge in some way, the idea being that some transformation or set of transformations will remove the perceived anomaly. Hence, the process is, like in Dervin’s model, a potentially iterative one.
This is where knowledge of how the user perceives an event becomes useful. Belkin proposes that in order to better serve a user, the system should focus its query language on the ability to express an ASK, and furthermore should base its document retrieval behavior based on the ASK of the user, not on a “best match” methodology employed by other systems Belkin, 140). Furthermore, he suggests the methodology of perhaps dividing ⚠ ASKs up into separate classes of questions that the system is able to understand and accurately respond to. What better a place to employ Dervin’s sense-making model?
return to Info310.HomePage