Skip to content

Latest commit

 

History

History
157 lines (82 loc) · 23.9 KB

A Patron's View of the Library's Online Catalog.md

File metadata and controls

157 lines (82 loc) · 23.9 KB

A Patron's View of the Library's Online Catalog

The Las Vegas Clark County Library District (lvccld.org) has recently migrated its online catalog to BiblioCommons. There's a lot to like about the new catalog -- there are some big improvements from the previous catalog -- but it doesn't go very far in satisfying my hopes and dreams, or, I suspect, those of many digital natives.

Public libraries are in a tricky spot, even a bit worse than bricks 'n clicks companies. Public libraries provide many valuable services from their 'bricks' locations, but they also maintain a web site. The problem for the web site is that the Internet moves so fast. One day you're stuck buying shoes online because you can't remember your shoe size, and the next day the app will figure it out from a photo of your foot. People's expectations aren't set by what's reasonable or possible, but what they've seen on other sites (or apps).

It will be clear, if you make it to the end of this note, that I don't actually want a better online catalog. I want a PBA, a Personal Bibliographic Assistant. I blunder around using various tools to navigate the bibliographic universe -- mostly books -- and it's irritating a lot of the time. These notes (mostly) aren't theoretical objections to the catalog, but places where the catalog doesn't work for me. What I want isn't necessarily all that reasonable, but I believe it would be wrong to discard it for that reason. The Internet has demonstrated that today's stability can quickly become tomorrow's disruption. What if Amazon decided it would like to be a public library?

License: CC BY-NC-SA 4.0

Graeme Williams
Las Vegas, NV
[email protected]
github.com/lagbolt

What am I doing?

Well, other than obtaining, reading and returning books, I'm keeping track of books that I might like to read. I'm grateful to Cossham for pointing out that among the "activities and processes associated with the bibliographic universe" is "finding things to read" [1]. Quite.

This includes finding what's available.

A Patron's View of Availability

A book might be available at my local library branch, or somewhere in the system. It might be available via inter-library loan. Or I could buy it from Amazon or the local Barnes & Noble.

It might be unavailable at the library because it hasn't been published, or it has been published and the library decided not to buy it. The local branch might own a copy which is borrowed, or other branches might own copies which are borrowed.

Some of these possibilities might seem like a distinction without a difference, but they make a difference to me. For example, if the local branch owns a copy, I might choose to wait until it becomes available. If it doesn't, I might as well submit a hold request and let the book come to me. If WorldCat tells me that no library within 500 miles owns a book, I might as well order it from Amazon.

Each patron might react differently to each of these possibilities depending on how much time and money they have, but each of the possibilities is genuinely different. We'll see that the catalog doesn't do a particularly good job distinguishing one case from another.

My workflow

I don't really "find" things to read, since the Internet provides a firehose of book information and suggestions. My problem is really to pull books out of the firehose and keep track of them. I subscribe to quite a few blogs and mailing lists which let me know about upcoming science fiction and fantasy books, so I often find out about books before they are published. I'd like a good way to keep track of them.

What I do at the moment is to keep a list in Microsoft OneNote with the author, title, series information and whatever information is provided about publication date. There's nothing good about this. I don't automatically find out when a book is published, I don't find out if the library orders it, I have no good way to interact with the library catalog entry (e.g., to put a hold on it), I have no good way to generate an inter-library load (ILL) request, and I have no good way to buy it if the library doesn't.

I could keep track in WorldCat or Amazon, but neither is much better. WorldCat will show me whether LVCCLD has a copy of a book, but it won't notify me if that changes, and current availability information is too many clicks away.

What I want is a way to add (or link) unpublished books to my library account (e.g., to a shelf or list) with an easy way to convert to a catalog entry if the book is added to the catalog. Ideally, I'd like to know when the book is published, so I'd know if the library has decided not to acquire the book [2]. And I'd like an easy way (i.e., without retyping any information) to submit an ILL request.

The frustrating thing about BiblioCommons is that you can add a URL to a list, but the list item is not "smart". If you add a link to a book in WorldCat, BiblioCommons doesn't process the fact that it's a link to a book with (e.g.) author and title information, not to mention a unique identifier.

The flip side to keeping track of unpublished books is keeping track of published books which the library doesn't have. I regularly see lists of science fiction and fantasy books like, "The Ten Best Dragons in Fantasy", or, "The Ten Smartest Spaceships in Science Fiction". If a book catches my eye, I could look it up in the library catalog, but if it's not there, I can't tell whether there's a typo in the title or whether the library doesn't have it. Google has the best query parsing and the best error correction, so that's where I start. Goodreads normally appears in the first page of results (particularly since Google know that's what I click on), so I go there. Goodreads has always had good series information and good information on other books by the same author, as well as reader ratings [3].

If the book's rating on Goodreads looks OK, I do a search in the library catalog. If the book is in the catalog, I can add it to my For Later shelf. If it isn't, I look it up in WorldCat to get an idea if ILL is feasible (although the rules and cost of ILL are opaque to me). In either case, I add it to a page in OneNote, including the WorldCat "permalink" in case I do want to submit an ILL request.

For me, the most frustrating time is when a book has just been published but doesn't appear in the library catalog. I don't particularly want to buy a book if the library is going to get a copy -- but I have no way of knowing this until and unless the book appears on order in the catalog.

So far, I have books listed in my For Later shelf, as well as two different pages in OneNote. Now read on ...

There are a few places where I bump into actual physical books. I occasionally go into the local Barnes & Noble to take a look at their Science Fiction and Fantasy section. If I find an interesting book, the easiest thing to do is to scan it into my LibraryThing account using my phone's camera and the book's UPC code. I could search for it in the library catalog but typing the title in by hand, I'd be working against the catalog's error correction and limited holdings.

Whenever I'm in the library, I look at the new books section, and again the easiest thing to do is to scan interesting books into my LibraryThing account. (I have different tags for books I find in different places.)

One odd thing about the Las Vegas library system is that they don't catalog most paperbacks. (I say odd only because I haven't come across this before.) Unlike the rest of the fiction, the paperbacks are shelved by genre, so I can browse the science fiction and fantasy section, and again scan any interesting any interesting books into LibraryThing.

This leaves me with books listed in my For Later shelf, two different pages in OneNote, and under several tags in LibraryThing. Obviously I'd prefer to have a single place to keep track of books, where I could also see availability information.

WorldCat has all the book entries I need, and it does allow you to define lists, but it doesn't surface availability information. It's two clicks from a list to local library availability information for each item on the list. This doesn't sound like much of a barrier, but it breaks a common use case: seeing a list of available books -- for example, which books I could borrow if I walked into the library right now.

The Las Vegas BiblioCommons instance doesn't have WorldCat's almost complete coverage, but it does do a few things right with availability information. You can choose one or more branches as preferred locations, and the availability information in search results, shelves and published lists will include whether an item is available at a preferred location.

Holdings versus the bibliographic universe

I don't think there's any argument that users live in the larger bibliographic universe rather than inside the holdings of their particular library (e.g., Coyle [4]). BiblioCommons is doing a few good things to expand what a patron can see. For example, you can see lists from patrons and staff at other libraries. You can search for lists from any library, and you can see any lists that a book belongs to. You can see comments from other libraries, and I presume that same is true of tags (which I can't verify because tags are anonymous).

Because you can see lists from other libraries, it's possible to see books that your library doesn't own -- although you can't search for them in the usual way. If you do come across one on a list, you can add it to your For Later shelf. Once it's on a shelf, you can also add it to a list. And although I haven't done it, it looks like you can comment or add tags to it.

Anyway, the extent of the searchable catalog isn't the same as the extent of the library's holdings: the searchable catalog includes items which the library doesn't hold, and the library holds items which are not included in the catalog.

The catalog includes entries from Overdrive for audiobooks which are available from Overdrive but the library hasn't purchased (e.g., item 1740550134). This item, by the way, seems to have a complete MARC record. I presume the library loaded MARC data from Overdrive without modification.

Contrariwise, the library "holds" items which are not included in the catalog. The library just announced that patrons have access to a collection of audiobooks at RBdigital -- but these audiobooks are not included in the catalog. Since audiobooks from Overdrive are included in the catalog, this means that patrons have to search in more than one place, run more than one app on their mobile devices, and can't have a complete "For Later" shelf.

Keeping track

Accepting that the online catalog will only give me a limited view of the bibliographic universe -- even the little piece of it that I'm interested in -- BiblioCommons has several different ways of collecting and labelling the books I'm interested in, including tags, shelves, and lists.

Before I dive into details, I want to remark on something that I found remarkable: apparently, BiblioCommons thinks you can tell people what to do. Here's BiblioCommons product marketing on how that worked out [5]:

   "Lists in BiblioCommons are meant to serve one primary purpose: providing curated recommendations that are shared in the platform. ... However, our analysis of how lists were being used showed that a percentage of users were creating lists as a method to track the items they were reading or were going to read. Our users were essentially creating Lists instead of using Shelves."

Quite. People have been opening paint cans with screwdrivers for thousands of years.

BiblioCommons includes seven kinds of lists for recommendations as well as "If You Liked" lists; three built-in shelves: For Later, In Progress and Completed; and four kinds of tags: genre, tone, theme or personal. You can also add "Similar Titles" to a book's detail page.

This kind of patron data can be private -- like personal tags or private notes -- or public, like the other three kinds of tags. The trick is that you can't search your own private data, either in the simple search box or advanced search.

You can see why some patrons might want to use personal lists to keep track of their reading. You can have as many lists as you like: books that Mum might like, suggestions for the neighborhood book club, and reference books on rocket building or quilt making. If you try to use shelves for this, you've only got the three pre-defined shelves and the semantics won't fit. If you read a book that Mum might like, do you add it to "For Later" or "Completed"?

The problem is that BiblioCommons has stacked the deck against lists. You can add a book to a shelf in one click, compared to about ten to add to a list. And you can filter a shelf (i.e., using "facets") but you can't filter a list. I come across books to read far faster than I read them, so my "to read pile" will probably grow without limit -- pages and pages of books. So while lists look more flexible, it doesn't seem like they scale to the size I need [6].

If everything I'm interested in is one shelf, how do I keep it organized? How do I keep track of availability across several branches and the system as a whole?

Public tags like genre tags are a fine way of keeping elf novels and gnome novels separate, except that there are several kinds of public tags. Having several kinds of tags would be harmless fun except they leak into the user interface. Public tags can appear as a filter on searches, but each kind of tag is separate -- which means there'll be duplication. If you do a search on "Scalzi", for example, "aliens" and "military" appear in both Genre and Theme tags; "humour" [sic] appears in both Genre and Tone. Which means I have a tricky problem if I want funny books about aliens [7].

I can't sort my For Later shelf by availability, so I need a way to add that somehow. Private tags don't seem like a good way of encoding availability, but availability isn't as dynamic as you might think. A book enters the catalog when it is ordered, it's unavailable while it works through its hold queue, and then it's available. This kind of pseudo-availability ("available unless borrowed") is good enough for my purposes. If I look at a page of ten pseudo-available science fiction and fantasy books, at least nine are likely to be actually available. If I add an "available" tag when a book first becomes available, I can filter my For Later shelf by the "available" tag and see all the available books.

I want to keep track of availability at several different local branches, so I'll need a private tag for each branch. And there's no way to filter on the absence of a private tag, so I'll need a tag for "unavailable" as well. I can keep these tags up to date by regularly going through the "unavailable" items on my For Later shelf and -- since I can see the availability information right there on the page -- updating the tags on the items that have become available. Whenever I add something to my For Later shelf, I just have to set the tags right the first time. Then, before I visit a branch, I just have to filter my For Later shelf by the right tag and print it out.

Hold requests

I have to confess that the hold request system makes me so frustrated that I don't often use it for the intended purpose. Before the Minuteman system had lists, and before Las Vegas rolled out BiblioCommons, I used my hold list as a sort of For Later shelf, submitting a hold and then immediately freezing / pausing it. I'll try to explain my frustration below.

Other than the web site itself, a hold request is one of the few times the library communicates with the patron electronically. But what is a hold request? A hold request is an automatic response to a change in availability. This raises the obvious questions:

(i) for which patron? (ii) for what item? (iii) how has the availability changed? (iv) what action should take place?

Both online catalogs I've looked at allow patrons to freeze or pause a hold, which means you hold your place in the queue but won't get the Item [8]. So the library's answer is (i) the first patron in the hold queue with a non-frozen hold, (ii) when any Item of the specified Manifestation (iii) is returned, (iv) the Item will be shipped to the patron's local branch and held for the patron.

I mentioned above that I make a distinction between copies held by my local branch and copies held by the system. The reason for this is that if I can, I want to avoid the cost of trucking books from branch to branch. Yes, I know that for a single book the cost is very small -- call it teaspooning, if you like. But this means that I want a more complicated hold system. It would be better if I could put a hold on a specific Item -- the one held by the local branch.

That's not enough because for some books I'm willing to read either the large print or the "small print" edition. BiblioCommons certainly knows about different editions of the same book (i.e., different Manifestations of the same Expression) because it can group search results by Expression. I can't submit a hold request at the Expression level, but I don't want to, because I'm only interested in particular Manifestations, not any Manifestation of the Expression. I want to be able to pick the Manifestations my hold applies to.

But my primary frustration is that if I submit a hold request, the book inevitably arrives when I'm in the middle of another library book. There are often other people waiting for the new book, so I have to put the first book aside and switch to the new arrival. What I want is much closer to the old Netflix queue, where Netflix would send you a DVD once you returned the current one. The answer I want to question (i) above is sometimes, "the first patron with a non-frozen hold that doesn't have another book borrowed".

It's at this point, when I have constructed a hypothetical and ridiculously complicated new system, that I begin to doubt that I understand what I want. What do I want? Certainly, I want more information about availability. I can imagine two more answers to question (iv). Instead of physically reserving a copy, the catalog could send me an email in response to a change in availability. Or the web site itself could show changes in availability. In both of these case, I'll only see a book become become available when its hold queue is exhausted, but let's suppose that's OK.

If I go to a web site that sells something (physical stuff as opposed to e-stuff), I might find that something I want is out of stock. A sensible web site will provide a field where I can put in my email address to be notified when it becomes available. I don't see any reason why the OPAC can't do the same thing, although I suppose it might be hard to scale to systems with a few million patrons.

Certainly, BiblioCommons can do a better job of surfacing availability on the web site. Bizarrely, my "dashboard" will show me six items on my For Later shelf that are currently available, but I can't filter or sort my For Later shelf by availability. And I have no idea whether those six items -- or any other items on my For Later shelf -- are newly available.

The search results page can be grouped by Expression, so it's easy to see whether the library has a book, e-book, audiobook, etc., and for each whether it is available at all, available at one of your "preferred locations", if you have it on hold and the size of the hold queue. But it also won't show what's newly available.

Other lists (such as holds, shelves or your own lists) can't be grouped by Expression, so although they show availability, they're a lot less helpful than the search results page.

A book's details page has some information about the Expression it belongs to. Let's look at The Ghost Brigades by John Scalzi. The library has two editions of the book (probably the paperback and the hardcover) and the e-book. The details page for the 2006 hardcover (item S134C1521353) gives a clickable link for "all formats and editions", which is great, but on the page itself it only shows the availability for two unavailable "formats and editions", skipping the available 2007 paperback. It seems particularly odd to offer someone an unavailable e-book as an alternative to the unavailable hardcover when the presumably identical paperback is available.

Also, this feature (i.e., "all formats and editions") depends on accurate cataloging. For example, the book and e-book of Soonish by Kelly Weinersmith aren't linked, perhaps because the e-book has an error in the subtitle. (I note without further comment that the book record is from OCLC and the e-book record from Sky River.) It's times like this that I wish for a big red "FIX THIS RECORD" button.

Conclusions

I first started thinking about my finding and borrowing workflow several years ago, when the online catalog I was using was much less capable than the current version of BiblioCommons. It was clear that the catalog alone wouldn't be enough, so I experimented with code to log in and scrape information from the catalog, my account in particular. A moderate amount of code generated a moderate improvement in usability, and clarified my thoughts on surfacing availability. My expectations -- my hopes -- are based on those experiments rather than anything I've seen in an existing catalog.

BiblioCommons is much better than the previous catalogs I've used. It allows patrons to add data in a number of interesting ways, and makes a small start on making WEMI useful to patrons, but availability information still seems like a second-class citizen. I can't see how the features come together into sensible patron workflows.

Notes

[1] Table 6-1 on page 161, "Models of the bibliographic universe", Cossham, Amanda Frances, Ph.D. Thesis, Monash University, 2017. The version I have is 23549602Cosshamthesisrevised.pdf. Not that I read a lot of Library Science, but this is the first document I've come across that reads like the author has actually met public library patrons.

[2] I can infer that the library isn't going to get a book if WorldCat shows many libraries with holdings, but not Las Vegas.

[3] For a long time, OPACs got lost finding books by the same author because of the point at the end of the author's name in the MARC record, but thankfully that seems to have been solved. Also, Bibliocommons seems to have error correction which might be good enough for my purposes -- although it's limited to holdings and not as user-friendly as the incremental search in Goodreads or Google. On the other hand, series information is still weak: for example, it's missing from https://lvccld.bibliocommons.com/item/show/1911978134 but appears in the corresponding entries at GoodReads and LibraryThing. Series information is also missing from https://lvccld.bibliocommons.com/item/show/2067868134 and https://lvccld.bibliocommons.com/item/show/2142809134 and yes, they were the next two I checked.

[4] "The setting of boundaries around the library collection is one of the tenets of library cataloging goals — to define exactly what the library does and does not have. Although such an inventory is clearly needed, it is a mistake to also assume that this inventory and its boundaries is what interests today’s information seeker." from Coyle, FRBR, Before and After, at kcoyle.net.

[5] From https://www.bibliocommons.com/list-launch, copied to https://drive.google.com/file/d/1e1lDZ2dwlwUbRrJkVnOju5E2Fx5KVtbj/view?usp=sharing

[6] Lists are explicitly designed for a small number of items, so scaling to dozens of items is out of scope.

[7] The query language in advanced search avoids this problem: "au:scalzi AND tg:humour AND tg:aliens", because tg refers to any public tag. But still.

[8] I've capitalized Work - Expression - Manifestation - Item when I want to refer to the technical meaning -- as I understand it. See, for example, https://en.wikipedia.org/wiki/Functional_Requirements_for_Bibliographic_Records.