Cataloging Matters no. 19: Library Catalogs and Information Architecture

Cataloging Matters no. 19:
Library Catalogs and Information Architecture

Hello everyone and welcome to Cataloging Matters, a series of podcasts about the future of libraries and cataloging, coming to you from the most beautiful, and the most romantic city in the world, Rome, Italy. My name is Jim Weinheimer.
In this episode, I want to discuss the field of Information Architecture. I’ll talk a little about what it is, some of my personal involvement in it, and finally make a few suggestions for the catalog based on the field.

In the last couple of episodes, [no. 17 Catalogs as Data and no. 18 Problems with Library Catalogs] I discussed at some length and provided several examples, why I believe the library catalog is broken, and why FRBR, RDA, and even linked data do not solve the problems or even address them. In this episode, I would like to suggest some improvements based on the field of Information Architecture. I’ll be referring to several examples you will find in the transcript.
I’ll begin with a high-profile example of a web disaster: pathfinder.com.
In the transcript you’ll find a sample page of the site from 1998. Pathfinder was created as an information portal for Time/Warner in 1994. Actually, it was a rather typical example of an early corporate main webpage, one that was focused on the organization of the corporation. For someone within Time/Warner, this page was a reflection more of internal politics than anything else: for the CEO and other high-ups, they could readily see the vastness of their holdings.
For the heads of the individual sections, such as Time, People, Money and the other individual parts of Time/Warner, this kind of page hinted at intense turf wars since all could see the relative importance of their own area within the company. In the sample webpage, we can see that Time was the winner along with People, Money, Fortune and the others listed at the top while the losers got sentenced to the drop-down menu: the part that says CNN and other sites. You did not want to be an “other site”. When you opened up the menu (you can’t do that here since it is only an image) you would find that Lifeis in there, along with Dr. Ruth Westheimer and Time for Kids and lots of “other sites”. Even Time could claim that it was a loser here. In their opinion the only winner on this site was Pathfinder and everybody else lost. Yet, at least everything here is on the top page. To be consigned to one of the pages below this was to be sentenced to oblivion. And you had to do everything you could to prevent that. I provide a link to an article from Salon that discussed the problems not long after Time/Warner shut down the project. http://www.salon.com/1999/05/04/pathfinder/
Time/Warner spent millions on the pathfinder.com site and it was a disaster. Why? For a whole bunch of reasons, but primarily, it turned out that regular people outside the company couldn’t find anything at all. In fact, the public hated the site. Everyone in pathfinder.com was let go and the project was killed. The very top page still exists however at http://www.pathfinder.comand I have a copy of it in the transcript.
I am glad it is still there because I see it as similar to what Dante describes in his Inferno, the words that he and Virgil see written above the Gate to Hell: Abandon all hope, ye who enter here. This page is a warning: don’t go this way.
There were lots of sites like this in the early days of the world wide web and it took awhile for people to figure out who their websites were really for. Many are still learning this same painful lesson today. Does a website really exist just to make the CEO happy, or to provide a political statement of relative power within a corporation? Should it be a version of an annual report? Or does it have some other purpose? Information Architecture (or as I will call it: IA) stepped into the breech and brought some needed outside ideas into the equation.
In Wikipedia, IA is defined as: “the art and science of organizing and labelling data including: websites, intranets, online communities, software, books and other mediums of information, to support usability.” In this podcast, I am concentrating on websites. Anyway, Wikipedia has a pretty good definition, but it contains a lot. For instance, “to support usability” for whom? For the people who use your site. Not the machines but the people who use your site. Who are they? While you may think you know who they are, you need to find out. And the more you know about them, the better you can make your site.
I’d like to relate a story from my own past. When I was at Princeton University, I had written a paper cataloging manual that I called the“Slavic Cataloging Manual”originally to be used by the cataloging team at Princeton University Library that worked on the Slavic language materials. (It is now hosted at Indiana but my original version at Princeton still exists) Just a few of us used it. That was my entire focus. When the web appeared, my supervisor and I started experimenting with putting the department’s cataloging documentation online, and it was natural for me to start by working with my manual so it was one of the earliest sections. One day, in 1994 or 1995, I was looking at a page in the manual that we didn’t need any more so I took it down. To my great astonishment, the next morning I found an email waiting for me from an angry cataloger in Australia complaining that I had taken that page down. They needed it and wanted me to put it back up! Australia! That blew me away!
I put the page back up (a momentous decision in retrospect!), and over time I came to realize that I didn’t have the slightest notion who were the real users of my cataloging manual. There wasn’t that much on the web back then—those were the very early days—and I received lots of very strange questions from non-librarians who had no idea what my manual was about. I was even threatened a couple of times by some—I’ll just say it: weirdos who got angry at my manual! I began to ask myself: Who was the manual for? Was it just for our team at Princeton and I should forget the other Slavic catalogers in the world such as the one in Australia? And what about all those other people?
So, I started thinking about the primary concerns of IA rather early.
The field of Information Architecture was developing during the 1990s and the American Society for Information Science (they hadn’t yet added “& Technology” to their name) sponsored the first major meeting of information architects at the Information Architecture Summit in 2000. I was there myself and became the first moderator of their email list SIG-IA. That was a crazy experience on a crazy list and I learned a lot. There were very few librarians involved on that list; mainly there were webmasters of business and government sites, including many of those wild Internet startups.
At that time, the technology of “search” was primitive and still pretty much in pre-Google days (Google didn’t really start going until 1999 or 2000), and the existing search engines produced results that were pretty bad. Even so, webmasters—and especially the new information architects—understood that no matter how good searching might become, searching alone could never show what was really available on a specific website. There needed to be something more: something that showed people where they were on a site, as well as giving some idea of what was available on the site as a whole. Clearly, pathfinder.com and similar disasters had shown that structuring everything according to a company’s organization chart could only lead to ruin.
Not only did a site need a structure that “the users” could understand (whoever those users were determined to be), that structure also had to be extremely obvious to people who used the site. If people had to go hunting for something, the hyperlinks of the WWW, browser bookmarks, and search engines made it just too easy to go someplace else.
When applying this to businesses, these considerations were matters of life and death. Businesses desperately wanted to keep people browsing their sites for as long as they could, just as they want to keep customers browsing the goods in their stores, so that there is a greater chance that somebody might actually spend some money.
An essential corollary to these considerations that I learned from my experiences with the Slavic Cataloging Manual, and as I put up the rest of the cataloging documentation at Princeton, was something that didn’t make sense to me at first. My initial attempt at putting my Slavic cataloging manual onto the web was to take the same file I used to print out the manual, and just add the simplest hypertext markup I could find: <pre> at the beginning of the text and </pre> at the end, or “preformat”. Doing that retains the basic format of the text—all of the carriage returns and tabs and sometimes the bolds and italics. Anybody can do that in just a few seconds and you have a webpage. It was easy; it was quick, and kind of exciting because it was my very first webpage, but I quickly discovered that even though it was the same file that I had used to print out the manual, and the printed form had worked pretty well, it had somehow become useless online. At first, I thought the problem could be solved by improving the way it looked, so I worked very hard learning HTML to make the online version look as much like the printed version as I could, but it didn’t make any difference. Nobody could or would work with it online. Including me.
I thought we were all just being backward and stupid and refused to accept any changes but I gradually came to realize something else was happening. It turned out that I had to rework those materials in all kinds of ways that were new to me because if I didn’t, everything remained printed documents that just happened to be on the web and if people wanted to use them, they would print everything out because that was how the original documentation had been designed to be used. This was something I saw happen over and over again with my own eyes, until I finally concluded that all of that work I had done had resulted in failure. I wanted people to use the documentation they found online; not just as a storage place where everyone could find the latest version of the file that they would then print out.
At this point, lots of questions arise naturally, such as: how do people use your site? You don’t know and you need to find out. You may know how you use the site (and maybe you don’t), but someone else may use your site in ways you cannot imagine and you must design the site for those people and not just for yourself. After you learn how people use your site, you should be able to figure out if people are missing anything important. So, if you discover that lots of people are looking at one page but you know they should also be looking at another one because it’s closely related, but people aren’t looking at that second page, then you need to do something.
How do people come to your site? Do they start at the very top page and click their way through gradually, working from the top to the bottom, following any navigation you have provided, or do they arrive from a search engine and start out somewhere deep inside? Does that change anything?
There are lots and lots more questions that need to be asked.
It should be clear that IA is a rather complex field, and it is still relatively new. My own opinion is that most of the time, the information architect should come from the outside so that he or she can view the site with completely unbiased eyes. Very rarely, an internal specialist may be best, such as with my highly specialized Slavic Cataloging Manual, which was designed only for catalogers who are fully trained and already know the languages involved. They just need a little guidance and specialized knowledge.
I want to list a few of the basic tools and methods used in Information Architecture:
  • taxonomies and hierarchies
  • social navigation
  • breadcrumbs
  • sitemaps and tables of contents
  • site indexes
  • site guides
  • search wizards
That’s not all of them, but a few. Library catalogs already have some of them. The purpose of this podcast is to figure out how IA could help library catalogs, and unfortunately I don’t have one of the very first steps: I don’t have research that shows how people search a library catalog today. It really hasn’t been done. There are a few articles I can rely on and I provide links to them but mostly they say that nobody understands much of anything about library catalogs.
There is also anecdotal evidence I have gathered from email lists, what I have been told, what I have observed personally and what I have watched myself do. Nevertheless, I am an “insider” and by definition I believe that makes me an inferior information architect for this purpose. I know too much. But, I’ll give it a try anyway.
In my previous podcast, I described how and why the library catalog was originally designed to work: like a printed dictionary. I’ll summarize. Our 19th-century predecessors had various options for designing their catalogs, but chose the example of a dictionary because it was a tool that everyone already knew how to use. They understood that this model had serious defects but the fact was they assumed that everyone who came into the library had already learned their “abcs” in school and knew how to use a dictionary. When someone knew how to use a dictionary, and you learned just a couple of other things: to look up names by surname-forename and ignore initial articles, you pretty much knew how to use a dictionary catalog. Yes, there were still all kinds of specific problems, such as looking for someone named just “John”, say “King John”. Well, there are lots of people named John who may be popes or kings or saints or monks. So, you had to know how such names were arranged. But these kinds of problems didn’t come up all the time and people could always find out if they needed. Therefore, very little was required of the patron. If our predecessors had chosen other methods, which existed, people would have had to learn a lot more before they could use the catalog. Yet, our predecessors chose what they believed to be the easiest method for their patrons.
I went on to point out that those methods no longer work on the web, and describe how the catalog is broken as a result of that change. I demonstrated how it won’t even work for someone like me who understands it all and wants to use it. Now we see why it doesn’t work: the information architect assumes that something designed for another technology (printed cards) must be re-worked, or re-architected—an appalling word, I know—to function in a web environment, just as I had to re-work my Slavic Cataloging Manual.
When reconsidering matters, I suggest we start right where our predecessors did: when searching for information today, what do people already know? What are they used to?
It is not a printed dictionary, that’s for sure, or even a printed encyclopedia. Today, the answer is obvious: Google and other sites on the web. People today probably have more experience with these resources than most earlier people ever had with printed dictionaries or encyclopedias. That is a fact whether anybody likes it or not. Once that fact is accepted—something I feel is absolutely indisputable—the task of “architecting usability” in a library catalog opens up and becomes quite different. The question turns from “how do we teach the public to use our catalogs in the best ways” to “how do we make our catalogs as easy to use as possible for people? And do it today, not in ten years from now.” That means we base it on what people already know, that is, the websites people use everyday.
When people think of searching, they think of … Google. When people need information they no longer go to the Funk & Wagnall’s or even to the library. They “google” it and when they do so in the movies, it always works. What do people experience when they search Google? The Google page itself is almost empty. There is only a text box with a handful of links. People type a few words into the text box, and results appear magically that they can click on. It’s very forgiving, correcting typos and sometimes even finding synonyms. For most of the time of Google’s existence, people have been unable to rearrange the results, and it is only in the last few years that they can rearrange the results, by time, sites with images, reading level, related searches, plus a few other choices.

Some of those choices just disappear. Poof! For instance, remember that strange “Wonder Wheel” of a few years back that the Google crowd claimed was so wonderful?

 http://blog.jweinheimer.net/wp-content/uploads/2013/04/wonder-wheel-example22-276×300.jpg

 

I have a link to a post that described how wonderful the “Wonder Wheel” was. http://www.googlewonderwheel.com/google-wonder-wheel-step-by-step. Well, the wonderful “Wonder Wheel” just disappeared without any explanation, although when pressed Google’s employees mentioned “maintenance problems”. It then reappeared later under the alarming name: the “Contextual Targeting Tool”. http://www.google.it/ads/innovations/ctt.htmlFortunately, it has nothing whatsoever to do with drone strikes, but rather, is now aimed at extending Google’s Adwords program, which is one of the major ways Google makes its money. Very interesting.
The “Timeline” disappeared too and I kind of liked that one. I added a link to a Google video that announced these functions and how great they were. http://www.youtube.com/watch?v=MtirDMfcOKEIt’s kind of interesting to see the exciting announcements of some of these tools that have now disappeared, and nobody seems to care: the Uncle Sam search and Google Squared, and now Google Reader and iGoogle. All of these disappeared with very little fuss from the public. I think libraries can learn something from this, that the public can be very accepting.
Relatively few people really understand how Google functions. While the public may have some rudimentary ideas, they are so basic that they only tend to enforce misperceptions, as I outlined in the previous podcast. Even so, people do not need to know very much if they just want it to work well enough to find some useful information. Therefore, it’s much like a car. The starter and differential in your car are very complex but when you start the car or turn a corner, you don’t need to know how anything works. They just work. That’s enough.
When Google doesn’t work, what do people do? They type in other words. They keep trying new words until they find something that makes them “happy” (whatever that means) or they give up and decide that “there’s nothing out there”.
This behavior sounds similar to how the library catalog was originally designed to work. People didn’t have to know much to get something fairly useful out of it and if they found nothing, while they were supposed to ask a librarian, many walked out thinking, “What a lousy library. They haven’t got ANYTHING on the War of the Northern Aggression!”
What can we conclude from this? People probably search the online library catalog in ways that are familiar to them, and this means: the ways they search Google. In fact, if they didn’t search it the same way, it should be surprising. I emphasize the onlinecatalog here because if people were confronted with a physical card catalog, their reaction would probably be different.
This makes sense of a recent observation. On Autocat, someone mentioned that during an “information literacy” session after discussing what the library catalog is, the library website, and everything else, they asked the students to find out: when is the library open on Sunday. Most typed “when is the library open on sunday” into the library catalog.
Why not? It makes perfect sense from their own standpoint. For them, they saw just another “magic box” where they could type in some words and see what comes back. If however, those same students were looking at a card catalog, I doubt if any of them would have made that same mistake by walking over to the “W” drawers and looking for “when is the library open on sunday”.
From my own experience, people know practically nothing of catalogs and while some may actually have sympathy for the concept of “authorized forms” the ideas of cross-references or subject heading strings are very strange to them. They understand authorized forms more as descriptors or even more, tags that they see in online databases and other sites. This also goes for the idea of being able to search specifically by authors, titles and subjects, which many do not understand. This all stems from the fact that they can’t do any of that in Google, and Google is their basic point of reference.
Based on these considerations, I shall assume that the user of the catalog will be any member of the public who knows nothing about how to search an online library catalog or even what a library catalog is; many may not even realize they are in a library catalog since they could have followed a link from anywhere. And I assume these people will relate to the catalog as they would to a Google search. What are some of the solutions that IA could come up with? Let’s discuss viewing individual records first.
Creating Context
I personally believe this is the most important task of IA and explains why my Slavic Cataloging Manual failed when I put it up the first time. It is important for someone on a site to be able to orient themselves: to know where they are and have some kind of idea of the greater whole. One of the most widely used and, I think, successful methods of IA is to show someone where they are on a site by using something called “breadcrumbs”. I am sure everyone has seen these, and they can be implemented in different ways. In the transcript, I have an example of breadcrumbs from Wired.com.
Here, not only do we see an article about AT&T, we see that it is in the Section “Opinion – Business & Enterprise – Tech Policy & Law”. This does not seem to be a strictly hierarchical relationship but are two facets of “Opinion”. Of course, each of these are links to those sections where you can find more articles on those topics. At the very top, we see a general navigation bar. As a result, someone can come to this page from anywhere and quickly know: they are on the Wired.com site; they are in the Opinion section; they see that there are related sections that they may find interesting; and they have an overall idea of what may be on the site as a whole. Very well done and highly efficient.
How is this handled in Wikipedia? We find various types of “breadcrumbs” there. I have a screenclip from Wikipedia that shows the breadcrumbs (they call them Categories) and some templates that offer more dynamic navigation:
http://en.wikipedia.org/wiki/Zydeco
These are great breadcrumbs but I think they would be even better if they were farther up the page. Could something like this work in a library catalog when someone is looking at a single record?
I have a screenclip of a record for a book about Zydeco music from the LC catalog. LC is currently in the process of changing catalogs. This is the older display but the new one doesn’t change anything for our purposes.

 

Let’s examine this in comparison with the article from Wired. What are the breadcrumbs we see here? We see that we can do a new search and have different functions in the system. We see we are in the LC catalog. But more important: does this display give us a sense of where we are on the site? Or what related information is available? I’ll talk about the subject heading “Zydeco music—History and criticism” in a moment. I don’t see any context similar to what we saw in Wired. What can we do?
As I discussed in the previous podcast, if someone sees a record for “Zydeco music” as here, the searcher needs to know that there is such a thing as a “set of other resources on Zydeco music”, and that there may be information within at least one other set of records labelled “Popular music – Lousiana”. In the card catalog, the first task was handled automatically when someone looked at the subject card for this book, since people could only see that card with the other cards on the same subject. Again, the second task I will discuss in a moment.

In the transcript, I reworked the LC record to include breadcrumbs, similar to what we saw in Wired. It’s very simple. All I did was put in the entire subject hierarchy by adding “Music – Popular music – Louisiana – Zydeco music – History and criticism”. I also followed Wikipedia, labeling it “Categories”

Any of those levels could be clicked on. I think people could learn a lot from navigating in this way. Probably, people would quickly realize that Popular music—Louisiana would be useful to examine. As they worked with these arrangements, they could see that there are several geographic choices, including “Southern States” as I mentioned before; they find there are different types of music and so on.
There are many possibilities to create really useful breadcrumbs and thereby help people to understand a bit of what the structure of the catalog can provide.
But, I confess that I hate the word breadcrumbs.
Visualization
So, to return to our example, when someone would click on “Popular music” or “Louisiana” or another part of the breadcrumbs, what should happen? To repeat what I described in the previous episode, in our current catalogs, for those rare searchers who actually click on the subject and browse “Zydeco music” alphabetically, they see something very strange: they see items about the word “Jew” in Lithuanian, or corporate bodies and places in Poland! I have a screen clip that shows this.

 

This is the opposite of useful. Even Charles Cutter along with his colleagues all knew of this problem and discussed it, realized that this arrangement wasn’t useful, but they had little choice with the tools that they had.
So what would be useful? Should people see a list of the bibliographic records with that heading or should they see something else? My own opinion is that people should be given a choice: to go directly to the records or to continue to visualize the structure of the catalog. How could that work?
To illustrate it, let’s examine something more complicated: how about the entire topic of Music? I have a link to the LC subject browse for music where we find the traditional list of narrower terms in alphabetical order,and then the beginning of the subdivisions of Music, also in alphabetical order.
Are there other ways to view this type of information? There is the visualization available in id.loc.gov but I find it absolutely terrifying. I have a link to the Music subject http://id.loc.gov/authorities/subjects/sh85088762.htmlbut beware! Click on the tab marked Visualization at your own risk!

Of course, there are other ways to do it. I have a link to the page “Outline of Music” in Wikipedia http://en.wikipedia.org/wiki/Outline_of_musicwhich is also pretty impressive. The page itself uses a classified arrangement (yet more evidence of eliminating alphabetical arrangement), but there is another view I would like to point out through the WikiMindMap.

This method provides a nice, non-threatening, cool, interactive view of the outline of Music we see in Wikipedia, including links into the articles themselves. If you haven’t seen it, look at it. The person can choose either to continue navigating the concepts in the MindMap or go directly into an article.

Relating this to our catalogs, it could work similarly, where you would have the choice to navigate the topics or go into the bibliographic records immediately. A mind map based on LCSH would of course look different from the one we see in Wikipedia: there would not be “People” or “Instruments” as we see here, but instead “Anger in music”, “Devil in music”, “Music for physical therapy” and so on. I personally believe that people would love this type of display for our headings.
There has been a lot of experimentation under way to display data, including data that is essentially hierarchical, which is what much of catalogs are. I provide a link to a list of projects http://www.wikiviz.org/wiki/Tools#Tools_for_network_data_and_hierarchical_dataSome mapping tools I do not care for but my personal opinions are irrelevant. What is important is whether the public likes something or not because the current visualization, alphabetical order, doesn’t work—it can’t work—and other methods exist that need to be tried.
Facets
I believe that facets are related to visualization. I’ve spoken a lot about them already. In short, I think facets represent a huge advance for catalogs, but I suspect that most people do not really understand them. This is probably the way they are displayed. Presenting the facets as I have normally seen them, in a left-side menu with everything truncated to display only the top five or so results of a specific facet doesn’t serve the searcher who needs to get an idea of what is available. I have a sample screenclip from Worldcat.
In a recent paper, I suggested an alternative method for display of facets. Reality Check: What is it that the Public Wants today? Strangely, it is now possible for the catalog itself to explain the results of a search in words—actual text that could even be converted into audio.
Search Wizards
When searching, much of the problem is to come up with the correct query. One way IA has solved this has been to create what are called Search Wizards, which actually create the search for you. There are many examples of this. I have a link to Samsung’s download site which consists of three drop-down menus: category, sub-category and model code. http://www.samsung.com/us/support/downloads/global
The choices change based on what you do, so if someone wants the manual for their Samsung DVD recorder, they choose “DVD Player” under categories and the sub-categories will then show options restricted to DVD players, e.g. “DVD Recorder” and from there they can find all the model numbers of DVD Recorders.
Another type of search wizard is Google Suggest, which occurs when you begin typing into the Google text box and a dropdown list appears with suggested terms. I provide a screen clip where I have typed in “zydeco music”.

By the way, people play with this, typing in “why do men …” or “why do women …” to see what comes up. It can be funny. Several catalogs already have this type of functionality, including Worldcat. It seems to me that combining the functions we see in Google and Worldcat that suggest terms, with the multi-box function we saw in Samsung could help people create some very complex queries using main topics and subdivisions.

Another example of a type of search wizard is Google Recipes, where if you enter some ingredients you happen to have in the house, e.g. chicken and lentils, https://www.google.com/search?num=100&hl=en&newwindow=1&safe=off&tbm=rcp&q=chicken+lentilsyou can get recipes with those ingredients, then refine it by other ingredients, cooking time, and so on. If you look at the example in Google I provide, you have to click on “Search Tools” to see how to refine the search.

Google Recipesis pretty amazing. How in the world does Google do it? Actually, Google doesn’t, but this is an example of how semantic technologies work. If somebody wants their recipe included for Google manipulation, they have to mark up their recipes in special ways that are very simple and almost anyone can do it. I have a link describing how to do it. http://support.google.com/webmasters/bin/answer.py?hl=en&answer=173379
This shows how things could work with catalog data. You don’t need full-fledged RDF—in fact, RDF won’t work with these recipes. You don’t even need linked data or the semantic web. You need to do things the way Google tells you to, and that is by using other, much simpler formats available from schema.org.
Social navigation
The main idea of social navigation is that the crowd can see a lot more than you can and therefore you should be able to use the power of the crowd to find better information or at least to find information more efficiently. Twitter is an example. The big talk right now is Graph Search from Facebook. https://www.facebook.com/about/graphsearchI haven’t been able to use it, but from the videos describing it, I confess I am skeptical if it will be useful for more serious information search and retrieval.

For me, the best example of social navigation is seen in Amazon.com. I include a screenclip of part of Amazon’s record of the book I used before on Zydeco music.

The breadcrumbs we see at the top are “New Releases” “Best Sellers” “Children’s Books” and so on that are not very relevant for someone looking at a book on the History of Zydeco music. More useful are the two links under “Customers also viewed these available items”. This is “Social navigation” in action because these items appear based on what others have done. Going down the page, we see all kinds of information, editorial and customer reviews and it is not until the very bottom that we find the bibliographic description of the item along with the topical breadcrumbs that are very broad. I’ll read them out very quickly.

 

These appear to be BISAC terms. There is no mention of Louisiana or Zydeco. By placing these topical breadcrumbs at the very bottom of the page, or in other words, the part of the page that the fewest number of people ever look at, Amazon shows that it is obviously relying on their social navigation tools much more than their breadcrumbs, which are not nearly as specific as the LC Subject Headings. While reliance on the “crowd” may be fine for a company that makes money selling its goods, I would hesitate to rely too much on such a method in a library catalog.
Currently there is limited use of such tools in our catalogs, although many offer tagging and comments, people can often make virtual bookshelves and share pages with Facebook, Twitter or the like. Of course, we could work with log files and circulation information to show what others have done, just as Amazon does. Although there are many areas where social navigation may be profitable for the public, I am focusing here on Information Architecture in the library catalog so I don’t know if there is much of use. I shall go out on a limb and make one suggestion: if we could place the authority files into a type of wiki so that people could annotate those records, add links and other information, perhaps a lot of good could come out of it. But that is just an idea.
Guides
Very few people will ever read a guide for how to use a site—have you ever read one?—but that doesn’t mean a guide should not exist and that guides cannot be useful, but they must be made in such ways so that they are useful. One real success in one of my catalogs were a series of what I called “Two-Minute Tutorials” where I created a series of guides where each could be read in a maximum of two minutes but many were much shorter. In the catalog, I tried to have links to the tutorials appear only when someone encountered a problem so people did not constantly see the same things and learn to ignore them.
Whenever I found that my patrons were having a problem, I could add a tutorial: how to use a specific bibliographic tool or how to revise a search. Sometimes there was just too much to cover in two minutes, for instance I wrote one entitled “Working with Controlled Vocabulary”. I could only try to make people aware of possibilities and some very basic methods, but I provided links into the more substantial guide that I had written that people could click on if they wanted. Sometimes they did.
I discovered that although I would get only a handful of real reference questions in a month, these tutorials would be accessed hundreds of times a month, and that was counting only accesses from Rome, or my own patrons. Therefore, I considered that a success.
Some methods of IA do not seem particularly suitable for a catalog, at least in my opinion, because IA mainly focuses on how web masters can create the best navigation for an individual website. That does make it a little different from a library catalog. Therefore, the IA staple of a sitemap or site indexes do not seem to be particularly appropriate, although we could say that the dictionary arrangement provides that now. I say this even though I made site indexes for the Princeton Cataloging Documentation and the Slavic Cataloging Manual. In alphabetical order with cross-references! But those were for individual websites. http://library.princeton.edu/departments/tsd/katmandu/genindex.html
http://www.indiana.edu/~libslav/slavcatman/slindex.html
Other major concerns for an Information Architect have already been solved in relation to the catalog, such as creating a thesaurus and creating a labelling system (authorized forms).
The Future
We are in a world where things are changing incredibly fast. Only a few years ago, everything was designed for a tabletop PC or a laptop. Today, everyone is designing for the mobile information world, the smartphones and tablets without keyboards or a mouse, but touchscreens, and perhaps in just a few years everyone may be wearing Google Glasses or some kind of clone of those spectacles and commands will be given through speech and body movements.
I can already search the web simply by speaking into my phone, just like in the original Star Trek series! This will have consequences for the Information Architecture of all sites, and will definitely make a difference in how people work with catalogs.
In the field of “search” Google’s advertising people are now pounding the drums to tell us how great is Google’s Knowledge Graph, which is supposed to show everyone the awesome power of the Semantic Web. Based on what I have seen so far, I think it’s a dud but who knows what they will come up with? http://www.google.com/insidesearch/features/search/knowledge.html
For me, the most interesting of these developments by far is what I have seen on the blog SearchReSearch by Google employee Daniel Russell where he shows the kinds of questions that are possible to answer today. The questions he asks and answers are completely different and could never have been done before. A typical example of one of these questions popped up recently. He has a picture of a yellow blotch on a crosswalk in Chicago and then he notices a similar one in Washington D.C. He asks: what is it? One of those nightmarish reference questions! But not only could he find an answer, he was able to do it in three minutes! It was easy! http://searchresearch1.blogspot.it/2013/01/wednesday-search-challenge-11613-whats.html
The fact that completely new kinds of questions can now be answered will have to have an effect on the public eventually. They will expect more and more.
I think that’s plenty on this subject for now. I could go on for a long time. In any case, those are some of the thoughts I have had on the subject that I think could make the catalog far more useful to the public than it is now, and certainly better than typing out the cataloging abbreviations! As I keep repeating, the first chore is to experience the catalog from the point of view of someone who knows nothing but wants to learn what your library has to offer—not what your catalog can do because they don’t really care, but what your library has to offer. Then we can begin to design something that will help those people, and will help them today.
The question of how to improve the catalog tomorrow is certainly an interesting discussion but essentially useless because changes are coming so quickly. We don’t know what people will want only five years from now. We don’t know what the technology will be. We don’t know what the economic climate will be. Youtube didn’t become really big until about 2007 or 2008. The first Ipad came out in 2010. That is the pace of change we are experiencing.
Information Architecture is based on helping people find what is on a site today, not what may be possible in ten years. It is not based on theory but on success and failure and as a result, any theory IA has comes from practice. IA is an evolutionary model focused on overcoming the hurdles found today, failing but still learning, with the idea to succeed wherever possible and to become better prepared to tackle the hurdles that will inevitably arise tomorrow.
For those who want to learn more about IA, I have some additional links at the end of the transcript.

The music to end this episode is the short madrigal “Chi la gagliarda” or “Who wants to learn the galliard” by Baldassare Donato, who worked in Venice at the end of the 16th century. This is performed by The De La Salle University Chorale of Manila. By the way, the galliard was a very energetic dance where you jump and it involved touching your partner. The man may even lift the woman in the air and many believed this dance was inappropriate. Apparently, the galliard was Queen Elizabeth’s favorite dance and was the way she kept in shape. I include a link to the text with a translation http://www3.cpdl.org/wiki/index.php/Chi_la_gagliarda_%28Baldassare_Donato%29 http://www.youtube.com/watch?v=zywxEFv0LYc
That’s it for now. Thank you for listening to Cataloging Matters with Jim Weinheimer, coming to you from Rome, Italy, the most beautiful, and the most romantic city in the world.

Selected Works on Information Architecture
If you would like to learn more about IA, the major text is:
Morville, Peter, and Louis Rosenfeld. 2006. Information architecture for the World Wide Web. Sebastopol, (Calif.) : O’Reilly.
I shall share the list of websites that I follow:
There’s even a course in iTunes U but it may be incomplete.
Finally, there are many videos on IA in YouTube.

-715

Share