Discussion Week 6
LIBR 202 – Section 18 – Dr. Bolin
“Online Subject Catalog Access”
Patricia Ayame Thomson
“Bates Suggestions to Improve Online Catalog Access”
For this week’s reading, I selected the article by Marcia Bates titled: “Subject Access in Online Catalogs: A Design Model.” Bates opens the article with a rather ambitious and grand statement as follows: “A model based on strikingly different philosophical assumptions from those currently popular is proposed for the design of online subject catalog access.” At first glance, I thought Bates was proposing a “strikingly different” technological design and reinvented a brand new online subject catalog access. However, upon closer inspection of the wording, what Bates is actually proposing is the following: “A model based on strikingly different philosophical assumptions from those currently popular.”
What Bates is proposing is not the latest, greatest, fastest, state-of-the art, new technological software program, but subtle changes in the philosophy to improve the current online subject catalog access programs. Albeit, philosophical and/or conceptual thoughts are not tangible, Bates argues that applying the concepts will aid in improving existing online subject catalog access.
“Optimistic Outlook About the Technology Revolution”
In her article, Bates shares an optimistic view about the technological revolution. Instead of looking at it as something to be feared or avoided, Bates welcomes it courageously and reinforces all the potential in the early days of technology. The article impresses upon us that it is the time to be pioneers and innovators in the relatively new field of the technological revolution. She points out that the current antiquated online catalogs are equivalent to the “horse and buggy” during the early 1900’s. Eventually, Bates’ overly-ambitious and optimistic ideal goes so far as to suggest that unless a database or online catalog is amusing and/or entertaining to the user, the system should be considered “defective.” However extreme Bates may sound, I believe the point she is trying to make is that anything is possible and there is still room for program design improvement and renovation.
On a brighter note, Bates says there has been some improvement in online catalog features and mentions: “We are just beginning to see features being designed for online catalogs which take fuller advantage of the capabilities of online systems.” Some of these features include: “natural-language search queries and automatic term stemming and weighting.” Also, Bates mentions the usefulness of “hypertext” in subject headings. “Hypertext” helps the user by actually guiding them to a different webpage that’s one step closer to their goal. In addition, the “hypertext” feature breaks up the linearity in a traditional file (like the one in physical card catalogs) and the record provides various links spreading out to reach many other sources.
Also, Bates argues the following point: “The proposed model is ‘wrapped around’ existing Library of Congress subject-heading indexing in such a way as to enhance access greatly without requiring reindexing.” The strength of Bates’ proposed model is that it does not require reindexing, and as a result, save an enormous cost reindexing all existing records. In truth, I believe the biggest difference in Bates’ model is the philosophical (mental) approach to searching, which can be very effective without costing a lot to integrate in the system.
The changes recommended by Bates do not reinvent the physical software program and requires some programming changes in pre-coordination and post-coordination. In Bates words, she states: “It is argued that both for cost reasons and in principle this is a superior approach to other design philosophies.” For the above reasons, Bates explains: “An end-user thesaurus and a front-end system mind are presented as examples of online catalog system components to improve searcher success during entry and orientation.” As a result, the low cost of Bates’ suggested minor changes in the front-end system mind (pre-co) and equip the user with an “indexer thesaurus” and/or “back-end thesaurus” (post-co) on the system. These supplementary system components suggested by Bates can be integrated in any current online system without altering the Library of Congress Subject Headings (LCSH) or compromising the main principles.
Bates presents three design principles in her article. They are: “uncertainty, variety, and complexity.” She states as follows: “Design features presented are an access phase, including entry, and orientation, a hunting phase, and a selection phase.” Specifically, Bates focuses on the beginning stages of the user’s search process in relationship to the search system. Below are Bates’ three design principles:
Principles
FYI: (Bates’ definitions are in Italics. I am paraphrasing Bates’ description of each principle underneath each definition.)
Uncertainty: Document description and query development are indeterminate and probabilistic beyond a certain point.
Bates explains there are too many unknown variables in the online search process to ever make it perfectly consistent and/or predictable. Based on a study, they found that even the same indexer tags the same item using two different terms. Attributes assigned are subjective and varies depending on the indexer, and as a result, can affect the rate of successful searches. At the same time, users who are the other half of the equation are also unknown variables. Every user enters a different word(s) (or natural language/ keywords) to search for the same information. As a result, the indexer must anticipate a large variety of possibilities in the queries. In the same way, in the design process the terms assigned for the controlled vocabulary can also be subjective and vary from indexer to indexer. If it seems like we’re going in a circle, that’s because we are. One is interdependent on the other, (ie: the indexer is interdependent with the user, and visa versa.) There is a large area of uncertainty in the online search process, since the input and as a result outcome will always be unpredictable.
Variety: Variety of query formulation must be as great as variety of document description for successful search.
In order to describe the Principle of “variety,” Bates quotes Ashby’s law as follows: “The law of requisite variety holds that for a system (whether machine or organism) to function successfully, it must generate as much variety in its responses to the environment as the environment generates as input to the system.” Bates suggestion includes helping the user become aware of the variety of words available. By showing the list of controlled vocabulary designed for the system, or a thesaurus for the end-user to provide more choices in term selection. As a result, Bates conveys that tools for users provided within the system will expand their knowledge about the extensive vocabulary available, which in turn will help them in the search process. Of course, the law of variety also applies to the indexer as well. The more variety the indexer/designer can come up with, the more it will apply to and match the search words (keywords) entered by users. As another tip for successful searches, Bates states the importance of convincing users to enter two to three terms together in the search box, instead of one at a time. The bottom line of the definition in this context is as follows: the more “variety” (provided from both user and indexer,) the more it will enhance the possibility of successful searches.
Complexity: Entry to and use of an information system is a complex and subtle process.
In the principle of “complexity,” Bates explains that the search and retrieval process is difficult and complex even in the human brain. Studies have found that every person associates certain words based on their own perceptions of life. When a person is asked to say a word that is associated to the one the researcher says, every person tested comes up with a different answer. The act of recalling information from memory is not a uniform and consistent function in the human mind, therefore it is safe to conclude that the online catalog system is equally as inconsistent and fickle. The ideal search engine’s attributes consists of: consistency and recall. In Chapter 3 of “Ambient Findability” by Moore (1995) states that as the documents in the system increases in number, the quality of recall drops. As the system grew, it “consistently” retrieved and dispensed results, but the “recall” became increasingly poor. Meaning, even though some documents were retrieved consistently, they were not relevant. It stands to reason that if the content of the results were not relevant, the search cannot be considered successful.
“In her article, Bates’ uses human analogies to convey machine functioning. I found this somewhat disconcerting.”
I find the way Bates describes her point using humans (or living organisms) as having analogous experiences with machines slightly disturbing. I feel there is a definite distinction between machines and humans. The most significant distinction is the following: Humans can think, feel, and evolve, and machines can’t. In my opinion, comparing a very human experience (such as grief) with a machine is far-fetched and inappropriate. Otherwise, Bates has legitimate and workable suggestions that can be applied to any online catalog systems in order to enhance search results.
“Subject Headings Not Viewed as Essential”
Bates refers to an article stating: “The possibilities that the subject approach might be important to students and that the relatively low use might be due to inadequate design of the subject access were never considered.” To further the point, she states in another occasion: “Rather, it appears that there has been something of a bias against library subject access; it has been seen as a nice extra, rather than as something essential.” I’m afraid I have to fall into the above category as well — before taking this class. I thought subject headings were frivolous functions (like decoration or something.) It never occurred to me that they may be “essential!” I never realize how much thought, work, and effort go into classification, cataloging, and indexing by the Library of Congress, for example. Little did I know that “subject headings” are inherently designed in the system to expedite and facilitate the user’s search. Now that I know why subject headings are there, I will definitely use them to narrow the search from now on.