The Catalogue Projects

The convergence of the Butterflies & Aliens Library coming into existence in January of 2021 and our Head Alien taking LIS 502: The Organization of Information in September of 2021 has resulted in the spin-off of two experimental projects to catalogue at least part of the actual book collections that fall under our virtual umbrella (and part of which is also serving as a final project for LIS 502… more below).

The key word here is experimental.

We are currently running two experiments in parallel, PROJECT COBOHOS and PROJECT INTERLOPER.

About The Catalogue Projects

Dec 7, 2021 • Written By Winston Pei

In describing these projects, a large proportion of this article is serving double duty as the essay component for my final project in LIS 502: The Organization of Information, one of five core courses for the Master of Library and Information Studies (MLIS) degree at the University of Alberta School of Library and Information Studies.

My official submission is the digital library of my collection of alphabet books, with all items fully reviewed and tagged. But the full scope of The Catalogue Projects has expanded well beyond that first collection.

In fact, these experiments began somewhat independent of being a class project at all. In reaction to early readings in our textbook… Joudrey, Taylor, and Wisser’s The Organization of Information (2018), especially Chapter 3: Retrieval Tools… I began exploring options for cataloguing my own books, “to help users find, identify, select, obtain, and explore information resources” (Joudrey, 2018, p. 97). When I discovered Libib.com, a cloud-based cataloguing platform with built-in OPAC and circulation management capabilities, and a remarkably fast barcode scanner built into its mobile app, cataloguing my existing alphabet book collection – literally “Starting with our ABCs…” – became a natural first test, now expanded to become PROJECT COBOHOS. Then exporting that data and importing it into a previously dormant LibraryThing account, and activating its OPAC module TinyCat, became the natural extension to the experiment as PROJECT INTERLOPER.

Which brings us to the final project for LIS 502…

Aside from my own personal use, the target users for these catalogues, and the Alphabet Book Collection in particular, are members of the Books & Bricks Irregulars, attendees to what in pre-pandemic times was a monthly in-person open house. With such events restricted by the pandemic, these catalogues allow for virtual browsing of surrogate records, with further access possible either by “patio pickup” lending or video show-and-tell at our now-virtual monthly events, with the added bonus of being able to extend Books & Bricks access to people who had previously been unable to attend because of timing or distance.

As a result of this initial driver, ‘collection’ became a primary facet of metadata in that it became, somewhat like a Dewey Decimal or Library of Congress number, a type of broad call number that combines both primary subject metadata and the physical location for the actual resource. As such, the exercise of naming and refining the ‘collection’ facet – basically a high-level folksonomy (Joudrey, 2018, p. 526) previously only loosely embodied in how I grouped my books across shelves and rooms – became an iterative process that involved physically moving around books as much as identifying and changing metadata tags.

From there it became an exercise in adjusting and adapting even basic metadata like title to fit constraints and limitations in both the Libib and LibraryThing platforms, for example how they alphabetize titles in different ways, which fields are searchable, and the lack of International Standard Book Number (ISBN) for older books. This included, for the Alphabet Collection in particular, contacting Libib support to figure how to bypass the system default of ignoring “a” when alphabetizing titles since the “a” in A is for Activist serves a much different function than the “a” in A Bad Kitty Christmas.

It was interesting to note just how much the software influenced metadata tag development. Libib’s simpler and visually more elegant OPAC interface better serves our target audience for the purposes of browsing, with keyword- and tag-based searches also available in limited form, so the focus was as much on how the metadata tags would display and function as what they were. Meanwhile LibraryThing functions with a philosophically more librarianship-based underpinning, having embedded within it ideas like the distinction between work, expression, manifestation, and item from Functional Requirements for Bibliographic Records (FRBR) guidelines (Joudrey, 2018, p. 294). It also incorporates several controlled vocabularies, alongside user-generated tags, for shared information like genre and format, much as you might find in any larger integrated library system (ILS). As a result, some migrated tags from Libib ended up being redundant.

At the same time, these respective features also came with weaknesses. It quickly became clear that the simpler interface in Libib has also meant greater limitations, most notably the lack of any true faceted search capability. Meanwhile LibraryThing excels at faceted search, but introduced a significant other challenge, perhaps characteristic of any ILS, of metadata from outside sources that are inaccurate or at least do not accurately reflect the items in our actual collection.

In regards to cover images in particular, and the catalogue metadata generally, both Libib and LibraryThing use ISBN, when available, to pull in data from a variety of sources. The founder of Libib, Javod Khalaj, has said that “the service uses multiple open data sources, including Google Books and Open Library [and] also draws on catalogs of university libraries that have given libib permission to access their databases” (Scardelli, 2015). Meanwhile, LibraryThing allows you to choose from a large selection of sources, including Amazon and local libraries. As such, our metadata ends up being a combination of open source images and content, as well as our own photos and information gleaned from our actual items.

Given all that, these projects have also become an active and ongoing iterative exercise in resource description, resulting in a) some manifestation-relevant information being stored in the Libib searchable description field, b) more item-specific information being stored in Libib notes, and c) the “faceted” description of resources in Libib using a subject tagging scheme derived via literary warrant (Joudrey, 2018, p. 493), in essence a systemizing of my personal interests and biases, hacked into Libib’s single tag field and subsequently mapped over to LibraryThing. A further exercise of mapping MARC and Dublin Core record categories over the default Libib structure (to keep a door open to hypothetical future interoperability), as well as testing how different fields were searched and displayed, also contributed to the iterative process of developing a metadata scheme to suit our catalogue and collections.

In both experiments, the platforms also differ internally in the functionality available on the OPAC side as opposed to the inward-facing library management side. One of the most problematic of these instances is the fact that tag-based search results in Libib default to a “by relevance” sort order and offers no other sort options, a noticeable issue when viewing only the Alphabet Book Collection as isolated via a tag-based search, although less so with The Catalogue Projects overall, where other access points and viewing/sorting options are available for the catalogue viewed in whole.

Going forward, certain existing tag-based facets in our PROJECT COBOHOS Libib experiment, such as genre, may well be eliminated as they are available through PROJECT INTERLOPER’s LibraryThing/TinyCat platform and its proper faceted searching capabilities, and are also arguably not a priority for my primary patron base. For them, a more limited set of Libib tags, each one a facet-plus-focus combination to ‘hack in’ a kind of faceted searching via the single tag field, combined with the much better visual browsing capability, should deliver an appropriate level of discovery functionality, very much akin to what they would have had when physically visiting our stacks.

– Respectfully submitted by Winston Pei in partial fulfilment of the requirements for Project 3 for LIS 502 with Dr. Lei Zhang, December 7, 2021.

Update: I got 90% on the project and a B+ in the class!

References

Dublin Core Metadata Initiative. (2020, January 20). DCMI Metadata Terms. https://www.dublincore.org/specifications/dublin-core/dcmi-terms/

Joudrey, D. N., Taylor, A. G., & Wisser, K. M. (2018). The organization of information (Fourth edition.). Libraries Unlimited, an imprint of ABC-CLIO, LLC.

Library of Congress. (n.d.) Library of Congress Classification Outline. https://www.loc.gov/catdir/cpso/lcco/

Library of Congress. (n.d.) MARC 21 Formats. https://www.loc.gov/marc/marcdocz.html

OCLC. (n.d.). DDC 23 Summaries. https://www.oclc.org/content/dam/oclc/dewey/ddc23-summaries.pdf

Scardilli, B. (2015, December 17). libib Helps Small Libraries Manage Their Collections. Information Today, Inc. http://newsbreaks.infotoday.com/NewsBreaks/libib-Helps-Small-Libraries-Manage-Their-Collections-108112.asp