Some thoughts on the state of free and open source software in libraries.

This article documents the steps taken in the preservation of a personal photo and video project, “Rural Route Nomad,” consisting of 14,058 born-digital objects from over a dozen different digital cameras used on world travels throughout all seven continents from the end of 2008 through 2009. Work was done independently, “DIY” if you will, with professional standards implemented in a manageable way sans the more extensive resources of a larger institution. Efforts were undertaken in three main stages: preservation, dataset generation, and visualization.

There are many methods to carry out a library book shift but those methods allow for varying degrees of predictability. The Bookshift.py script, when used in conjunction with accurate measurements of a library’s collection and shelving, provides library staff with predictability, flexibility, and the ability to shift in parallel. For every shelf, the script outputs a phrase such as the following, “The last book from this shelf goes 12.3 in/cm into shelf 776.” While complicated shifts can still create surprises, using Bookshift.py or similar methods typically make those surprises easy to correct.

While there are those within libraries that are able to take vendor APIs and use them to power applications and innovative workflows in their respective systems, there are those of us that may have heard of APIs but have only the slightest idea of how to actually make use of them. Often colleagues in various forums will mention that a task could be “just done with the API” but provide little information to take us from “this is what an API is” or “here’s the API documentation” to actually putting them to use. Looking for a way to automate tasks in Alma, the authors of this article both found themselves in such a position and then discovered Postman, an API platform with a user-friendly interface that simplifies sending API calls as well as using bulk and chained requests. This article gives a basic primer in how to set up Postman, how to use it to work with ExLibris’ Alma APIs, as well as the authors’ use cases in working with electronic inventory and course reserves.

SWITCH is a journal of new media art that has been published in an online-only format since 1995 by the CADRE Laboratory for New Media at San José State University (SJSU). The journal is distinctive in its commitment to presenting scholarship and criticism on new media art in a visual format that reflects and enhances its engagement with the subject. This approach, which includes the practice of redesigning the journal’s platform and visual presentation for each issue, raises significant challenges for the long-term preservation of the journal, as well as immediate issues related to indexing and discovery. This article describes the initial stages of a collaboration between the Martin Luther King, Jr. Library and the CADRE Laboratory at SJSU to archive and index SWITCH and to host a copy of the journal on SJSU’s institutional repository, SJSU ScholarWorks. It will describe the process of harvesting the journal, share scripts used to extract metadata and modify files to address accessibility and encoding issues, and discuss an ongoing curricular project that engages CADRE students in the process of augmenting metadata for SWITCH articles. The process reflects the challenges of creating an authentic version of this journal that is also discoverable and citable within the broader scholarly communication environment. This effort is part of a growing multi-institutional project to archive the new media art community in the Bay Area in a 3D web exhibition format.

Automated text classification is an important function for many AI systems relevant to libraries, including automated subject indexing and classification. When implemented using the traditional natural language processing (NLP) paradigm, one key part of the process is the normalization of words using stemming or lemmatization, which reduces the amount of linguistic variation and often improves the quality of classification. In this paper, we compare the output of seven different text lemmatization algorithms as well as two baseline methods. We measure how the choice of method affects the quality of text classification using example corpora in three languages. The experiments have been performed using the open source Annif toolkit for automated subject indexing and classification, but should generalize also to other NLP toolkits and similar text classification tasks. The results show that lemmatization methods in most cases outperform baseline methods in text classification particularly for Finnish and Swedish text, but not English, where baseline methods are most effective. The differences between lemmatization methods are quite small. The systematic comparison will help optimize text classification pipelines and inform the further development of the Annif toolkit to incorporate a wider choice of normalization methods.

Natural Language Processing (NLP) is a branch of Artificial Intelligence (AI) concerned with solving language tasks by modeling large amounts of textual data. Some NLP techniques use word embeddings which are semantic models where machine learning (ML) is used to learn to cluster semantically related words by learning about word co-occurrences in the original training text. Unfortunately, these models tend to reflect or even exaggerate biases that are present in the training corpus. Here we describe the Word Embedding Navigator (WEN), which is a tool for exploring word embedding models. We examine a specific potential use case for this tool: interactive discovery and neutralization of gender bias in word embedding models, and compare this human-in-the-loop approach to reducing bias in word embeddings with a debiasing post-processing technique.

Medium of performance—instruments, voices, and devices—is a frequent starting point in library users’ search for music resources. However, content and encoding standards for library cataloging have not been developed in a way that enables clear and consistent recording of medium of performance information. Consequently, unless specially configured, library discovery systems do not display medium of performance or provide this access point. Despite efforts to address this issue in the past decade in RDA, MARC, and the linked data environment, medium of performance information continues to be imprecise, dispersed across multiple fields or properties, and implied in other data elements. This article proposes revised definitions for “part,” “medium,” “performer,” and “ensemble,” along with a linked data model, the Ontology for Voice, Instruments, and Ensembles (OnVIE), that captures precise and complete medium of performance data reflecting music compositional practices, performance practices, and publishing conventions. The result is an independent medium of performance framework for recording searchable and machine-actionable metadata that can be hooked on to established library metadata ontologies and is widely applicable to printed and recorded classical, popular, jazz, and folk music. The clarity, simplicity, and extensibility of this model enable machine parsing so that the data can be searched, filtered, sorted, and displayed in multiple, creative ways.

This article will provide a brief overview of considerations made by the UTSC Library in selecting a persistent identifier scheme for digital collections in a mid-sized Canadian library.  ARKs were selected for their early support of digital object management, the low-cost, decentralized capabilities of the ARK system, and the usefulness of ARK URLs during system migration projects.  In the absence of a subscription to a centralized resolver service for ARKs, the UTSC Library Digital Scholarship Unit built an open source PHP-based application for minting, binding, managing, and tracking ARK IDs. This article will introduce the application’s architecture and affordances, which may be useful to others in the library community with similar use cases, as well as the approach to using ARKs planned for an Islandora 2.x system.

This article introduces CyprusArk, a work-in-progress solution to the problems that small museums in Cyprus have in providing online access to their collections. CyprusArk is an open-source web content management system for small museums’ online collections. Developed as part of Avgousti’s Ph.D. thesis, based on qualitative data collected from six small museums in Cyprus.