3.2 ASK for Information Retrieval: Part I Background and Theory
N.J. Belkin, R.N. Oddy and
This article discusses the background and theory of information need (IR) systems in relation to the “anomalous state of knowledge”. This state means that there is an anomaly in one’s knowledge that causes the information need. Belkin et al. discuss the idea that IR systems should be oriented as such that they can connect the information request to the information to solve the anomaly. The early IR systems used the idea of “best-match” which did not often solve the information need, especially if the need was poorly expressed. Like much of the literature we have read, this piece calls for a more user oriented system.
The biggest flaws I can see within the best match system assume that the user knows exactly what he or she wants, and can state it perfectly. The average person, even after using a particular system many times, may still have to try multiple queries before getting the desired result. In the best match situation, unless you only want information on the specific thing you ask for, you will not obtain as much information as you want. I know that if I want information on one topic that does not have many results; I will sometimes search another, related topic. Through a best match system it can be harder to get results.
Belkin and friends once again prove that a shift towards more user oriented systems need to be developed. While I’m sure there has been improvement since 1982, I can see some of their problems still within the information systems used today. Users often become very frustrated in trying to utilize many systems. I encounter this myself often at my job. I work with an online database and archive, and I often have to use a search function. However, the way this search function is designed, if you don’t enter works in the exact order they appear in the document, the document will not be found. Even two words searched but out of order, nothing will be found. I have to know exactly what I am looking for in order to find anything. It can be very frustrating, but I still have to use the database. In a library situation, the frustration could cause a withdrawal from using the automated systems. While librarians can be very useful, I feel like the goal is to get patrons to be able to use the systems effectively.
When reading the actual structure of the ASK system, I wonder if there are any databases used by libraries that function on an approximation of this system. As I know little to nothing about programming, I don’t know how possible or impossible this would be. If there is a system that uses something similar, does it work as well as Belkin and friends seem to think it will? Does it solve the problems it hopes to solve?
No comments:
Post a Comment