Large Language Models (LLMs) are increasingly used as flexible access layers for information-rich collections, raising the question of whether they can also support structured access to cultural heritage data as if they were digital libraries. This work investigates that question through Galois, a database-inspired system that treats the LLM as a noisy storage layer. We evaluate six querying modes: direct natural language prompting, direct SQL prompting, and four Galois strategies that combine LLM-facing tuple retrieval with deterministic relational processing. The evaluation uses two cultural heritage benchmarks: Paintings Extended, based on paintings data, and a new benchmark based on Europeana metadata. For both benchmarks, we define fourteen query templates with ten variants each. The main experiments use GPT-4o-mini, Granite 3.3 8B, and Llama 3 70B, with an additional diagnostic analysis of two GPT-5 deployments. We complement the benchmark metrics with a stratified manual audit of 60 unique disagreements. Performance is strongly shaped by model capability, deployment behavior, dataset structure, and execution strategy. Direct natural language prompting is the strongest and most robust baseline, while direct SQL prompting becomes nearly equivalent on the strongest open models. Among the structured strategies, only GaloisA and GaloisF become meaningfully competitive, especially on the flatter Europeana benchmark. In the audited sample, 35 disagreements were model-originated failures, 21 reflected reference-coverage or metadata-normalization effects, and four remained indeterminate. The additional GPT-5 deployment analysis reveals frequent empty outputs and provider-level constraints. These diagnostic results characterize the specific tested deployments and must not be interpreted as a general ranking of GPT-5 model capability. Querying LLMs as digital libraries is feasible, but its effectiveness depends on model strength, deployment conditions, dataset structure, and execution strategy. Direct prompting is already strong for exploratory cultural heritage access, while Galois remains valuable when relational discipline and controlled query execution are required.
At its core, a repository migration is the transfer of records from one system to another. In some cases, it is more of an unraveling, repair, relocation, and rebuilding of a historic property. In 2024, the International Food Policy Research Institute (IFPRI) library undertook its second migration from an independent CONTENTdm repository to a shared DSpace repository. The institute set an ambitious goal to complete the migration within a single contract cycle with CONTENTdm. Since 2012, the repository has functioned as a bibliographic metadata layer serving the main corporate website, numerous project sites, and the global consortium (CGIAR) website, as well as a digital publishing platform and institutional metrics and reporting system. The team started working with a combination of tools, including: Microsoft Excel, OpenRefine, and custom scripts. Microsoft Excel allowed staff to easily see and manipulate large amounts of data across different metadata fields. OpenRefine (version 3.7) was instrumental in normalizing and cleaning name and keyword variations, as well as locating and editing a single repeated datapoint. The scripts, created by a member of the team and author of this paper, saved hours of time and allowed the team to meet ambitious goals. The team successfully migrated over 20,000 records within 13 months while simultaneously improving metadata quality. Prior to the migration, about half of journal articles lacked external links and several dozen PDFs had been lost over time. Additionally, from 2008–2017, IFPRI’s records experienced an average of 40
Terminology and entity annotation tasks in digital-library and corpus-oriented workflows often involve expressions that are discontinuous, overlapping, or partially shared. While several existing annotation environments support span-based and relation-based annotation, discontinuous structures are frequently represented indirectly through workaround mechanisms, such as linking independent spans via auxiliary relations. This may increase annotation complexity and reduce the transparency of the resulting annotation objects. This paper introduces Annotaterm, a lightweight web-based annotation environment designed to support the direct annotation of contiguous, discontinuous, and overlapping terms through a fragment-based interaction model. This model supports simple, composite, and overlapping annotations in a uniform and inspectable way. A preliminary user evaluation with 14 participants suggests that the tool is perceived as suitable for terminology-oriented annotation tasks, particularly for handling overlapping and discontinuous expressions, while also identifying areas for future refinement. The proposed approach contributes both a conceptual model and a practical annotation environment for digital-library and terminology-oriented workflows requiring transparent handling of complex textual structures.
Conversational search systems enable natural, coherent dialogues between users and search systems to satisfy the user information need. The best of such systems improve user experience by asking clarifying questions (CQs) to resolve ambiguity in user queries. However, generating CQs that are both relevant and conversationally appropriate remains a significant challenge. To this end, we use Google’s “People Also Ask” (PAA) feature, grounding our questions in real-life user search behavior. We identify the most suitable methodology for this task by conducting an extensive comparative analysis of two methods: Parameter-Efficient Fine-Tuning (PEFT) and prompt engineering, and evaluate these using three state-of-the-art, lightweight language models. Furthermore, we introduce two new, human-annotated test sets derived from various conversational search datasets. Our results demonstrate that PEFT consistently and significantly outperforms all prompt engineering approaches across all models and test sets, and is statistically comparable to traditional full fine-tuning. These findings suggest that for the PAA-to-CQ task, investing in a high-quality dataset for efficient fine-tuning is a more reliable path to achieving high-quality, stylistically consistent outputs than relying on in-context learning alone.
This study examines the implementation of Freedom of Information (FoI) laws in Thailand and Indonesia, with particular emphasis on the relationship between FoI legislation and public sector records management practices. Employing a qualitative research methodology, the study applies the Records Continuum Model (RCM) as its principal analytical framework to evaluate the role of records in supporting transparency, accountability, and access to public information. Data were collected through documentary research involving the analysis of legislation, regulatory frameworks, and scholarly literature relating to FoI, records and archives management, and public information governance. A comparative case study approach was further utilised to identify similarities and differences in FoI implementation and records management capacity between the two countries. The findings indicate that FoI legislation functions not only as a legal mechanism for information access but also as an important framework for strengthening democratic governance, institutional accountability, and records management systems. However, both Thailand and Indonesia continue to face structural, organisational, and cultural challenges that constrain the effective implementation of FoI principles. The study demonstrates that the effectiveness of FoI laws depends significantly on the existence of robust records management systems, professional information governance structures, trained personnel, public awareness, and institutional commitment to transparency. From the perspective of the Records Continuum Model, public records are increasingly repositioned as shared societal assets that facilitate civic participation, public scrutiny, and democratic oversight. The study concludes that sustained institutional reform, digital infrastructure development, and professional education are essential for strengthening FoI implementation in the digital governance era.
The title of a research paper conveys its primary idea and, occasionally, its conclusions in a clear and concise manner. Choosing an appropriate title is often challenging, and automated title generation can assist authors in this task. In this work, we propose a technique to generate paper titles from abstracts using open-weight pre-trained and large language models. We use the CSPubSum and LREC-COLING-2024 datasets and introduce a new dataset, SpringerSSAT, curated from four Springer journals in the social sciences. Additionally, we use GPT-3.5-turbo in a zero-shot setting to generate titles. Model performance is evaluated with ROUGE, METEOR, MoverScore, BERTScore, and SciBERTScore metrics. Our experiments show that fine-tuned PEGASUS-large outperforms other models, including fine-tuned LLaMA-3-8B and zero-shot GPT-3.5-turbo, across most metrics. We further demonstrate that ChatGPT can generate creative paper titles. Overall, AI-generated titles are generally appropriate and reliable.
This study investigates the design and functionality of university library login pages across regional academic alliances (IVY Plus, BTAA, JULAC, JVU). By analyzing the login interfaces of member institutions, the research explores how the interfaces, their functionalities, security policies align with the collaboration and developments and also their reciprocal influences. The study aims to provide a general design checklist, which not only helps the UI design, but also uncover strategic patterns, policy-driven variations, and cross-regional trends that reflect shifts in academic library governance and digital access concerns. A multi-method approach was employed: screenshots and HTML files from 46 institutions were analyzed through categorization, statistical analysis, and comparative evaluation. Features were grouped into authentication mechanisms, usability, security/compliance, and library-specific elements. Core functionalities (e.g., ID/password, privacy policies) were consistent across alliances. Divergences emerged in feature emphasis: mature alliances (e.g., BTAA) prioritized resource accessibility with streamlined interfaces, while emerging consortia (e.g., JVU) emphasized cybersecurity (IP restrictions, third-party integrations). Usability features, particularly multilingual support, drove cross-alliance differences. ANOVA results highlighted regional and institutional influences, with older alliances favoring simplicity and newer ones adopting security-centric designs. This is the first systematic comparison of login page designs across academic alliances, offering insights into how regional, technological, and institutional factors shape digital resource access. Findings inform best practices for balancing security, usability, and accessibility in library interfaces.
Digital libraries are undergoing a profound transformation driven by advances in user experience (UX) design, artificial intelligence, and interactive technologies. No longer conceived solely as repositories for information retrieval, contemporary digital libraries increasingly support exploratory search, learning, collaboration, and sensemaking across diverse user populations. This special issue brings together research that examines next-generation user experiences in digital libraries, with a focus on novel interface paradigms, emerging interaction modalities, and inclusive, user-centred design methodologies. Collectively, the contributions highlight how UX innovation can enhance usability, accessibility, engagement, and learning, while also addressing the evolving expectations of scholars, students, and the wider public. The issue provides both conceptual frameworks and empirical evidence to guide researchers and practitioners in designing digital library systems that are adaptive, reflective, and responsive to human needs.
This review maps the intellectual base and evolution of Search as Learning using a reproducible workflow that combines a Web of Science corpus, Python screening, and VOSviewer, yielding 95 studies. Keyword co-occurrence, temporal overlays, co-citation, and country and journal mappings reveal eight thematic clusters and a shift from cognitive and exploratory themes to measurable learning outcomes and emerging AI mediated support. The co-citation results indicate a three part foundation: conceptual framing, operational measurement and prediction of knowledge gain, and log based behavioral methods. Collaboration concentrates in the United States and Northwestern Europe, with outlets centered on Conference on Human Information Interaction and Retrieval (CHIIR) and strong citation impact for the Journal of Information Science. Building on these findings, the review articulates an extended definition that unites user centered, interaction centered, and system centered perspectives, and proposes a simple prompt framework that operationalizes the system centered view in generative AI based SAL support.
Research contributions convey the essence of academic papers, highlighting the novel knowledge and understanding they provide compared to prior research. In this study, we address the challenge of extracting research contribution patterns automatically from citation sentences by proposing a tool-augmented LLM workflow, a novel framework that combines the precision of machine reading comprehension (MRC) with the contextual understanding and generative capabilities of LLM. The workflow operates through three stages: (1) Citation Marker Identification, where LLM locate and extract citation markers; (2) Contribution Patterns Extraction, which uses MRC to generate context-specific queries and extract INNOVATION, INFLUENCE, and FIELD patterns; and (3) Contribution Summarization, where the extracted triples are synthesized into standardized, fluent academic statements while correcting subtle extraction errors by LLM. Experimental results demonstrate that the proposed workflow outperforms generative baselines, such as ChatGPT, by maintaining structural consistency, avoiding factual drift, and preserving per-reference granularity in multi-citation contexts. The modular, stage-wise architecture enhances interpretability and facilitates error recovery, making it well-suited for real-world scholarly applications. In particular, the MRC-based extraction module achieves substantial improvements over the state-of-the-art W2NER model, with gains of +23.76 label-level F1-score and +31.92 entity-level F1-score. This work provides a scalable, transparent, and high-precision solution for automated research contribution mining from academic literature.
The timespan over which exploratory searching can occur, as well as the scope and volume of the search activities undertaken, can make it difficult for searchers to remember key details about their search activities. These difficulties are present both in the midst of searching as well as when resuming a search that spans multiple sessions. In this paper, we present a search interface design and prototype implementation to support cross-session exploratory search in a public digital library context. Search Timelines provides a visualization of current and past search activities via a dynamic timeline of the search activity (queries and saved resources). This timeline is presented at two levels of detail. An Overview Timeline is provided alongside the search results in a typical search engine results page design. A Detailed Timeline is provided in the workspace, where searchers can review the history of their search activities and their saved resources. A controlled laboratory study (n=32) was conducted to compare this approach to a baseline interface modelled after a typical public digital library search/workspace interface. Participants who used Search Timelines reported higher levels of user engagement, usability, and perceived knowledge gain, during an initial search session and when resuming the search after a 7-8 day interval. This came at the expense of the searchers taking more time to complete the search task, which we view as positive evidence of immersion in the cross-session exploratory search processes. Search Timelines serves as an example of how lightweight visualization approaches can be used to enhance typical search interface designs to support exploratory search. The results highlight the value of providing persistent representations of past search activities within the search interface.
This study investigates the emerging concept of “phygital” (physical and digital) user experience (UX) research within the context of public and academic libraries. It addresses two central questions: what considerations UX researchers and practitioners should keep in mind when studying phygital user interactions, and to what extent established UX research methods can be applied in these environments. Through a comprehensive literature review of English-language sources from the past decade across library and information science (LIS), human–computer interaction (HCI), and marketing, the authors examine the applicability of established UX research methods to phygital contexts. The study highlights several key considerations for library UX professionals, including the need to adapt methodologies, incorporate accessibility and inclusion frameworks, and navigate organizational challenges. The findings suggest that while existing UX literature offers valuable guidance, interdisciplinary collaboration drawing from fields such as marketing, HCI, and design justice can further support libraries in developing innovative and inclusive phygital user experiences.
Detection of concept drift in time-varying short text streams has numerous challenges since the data are volatile. According to research, 30
Digital libraries face challenges in quality, accessibility, and usage of resources. This issue presents seven papers offering computational and technical solutions to these problems: data quality through validation and monitoring, AI evaluation of information systems, and enhanced content discoverability. Research also covers knowledge representation with new provenance models, deep learning for bibliographic control, metadata-driven access to underrepresented languages, and computational methods for restoring historical documents. These papers showcase how modern techniques like machine learning, semantic web technologies, knowledge graphs, and image processing tackle digital library challenges, improving resource quality and accessibility. These papers were selected from the 21st Italian Research Conference on Digital Libraries (IRCDL 2025), held in Udine, Italy, on 20–21 February 2025, which has served since 2005 as a key annual forum bringing together researchers from academia, government, and industry to address topics spanning computer science, digital humanities, information science, librarianship, archival science, museum studies, and cultural heritage.
Nanopublishing is a paradigm enabling the representation of scientific claims in a distinctive, identifiable, citable, and reusable format, i.e., as a named graph. This approach can be applied to sentences extracted from scientific publications or triples within a Knowledge Base (KB). This way, one can track the provenance of assertions derived from a specific publication or database. However, nanopublications do not natively support multi-source scientific claims generated by aggregating different bodies of knowledge. This work extends the nanopublication model with knowledge provenance, capturing provenance information for assertions derived by an aggregation algorithm or a truth discovery process , e.g., an information extraction system aggregating several sources of knowledge to populate a Knowledge Base (KB). In these cases, provenance information cannot be attributed to a single source, but it is the result of an ensemble of evidence, that can comprehend supporting and conflicting pieces of evidence and truth values. Knowledge provenance is represented as a named graph following the PROV-K ontology, developed for the case. To show how knowledge provenance applies to a real-world scenario, we serialized gene expression-cancer associations generated by the Collaborative Oriented Relation Extraction (CORE) System. To demonstrate the value of trust relationships, we present a use case leveraging an existing scientific KB to construct a trust network employing three Large Language Model (LLM) agents. We analyzed the ability of LLMs to evaluate trustworthiness, exploiting techniques from KB accuracy estimation. We published 197, 511 assertions generated by the CORE system in the form of extended nanopublications with knowledge provenance. PROV-K also defines trust relationships between agents or between an agent and a proposition. Starting from these assertions, we leveraged external agents – namely, multiple LLMs – to assess their trusted truth value. Based on these values, we defined trust relationships between the agents and the facts, yielding an exemplar trust network comprising over 45,000 facts and four agents. The knowledge provenance graph allows the tracking of provenance for each piece of evidence contributing to the support or refutation of an assertion. To capture the semantics of the newly presented graph, we define the PROV-K ontology, designed to represent provenance information for multi-source assertions. The two use cases serve as a template to show how to serialize extended nanopublications and showcase the trust relationships’ capabilities.
Historical documents from Late Antiquity to the early Middle Ages often suffer from degraded image quality due to aging, inadequate preservation, and environmental factors, presenting significant challenges for paleographical analysis. These documents contain crucial graphical symbols representing administrative, economic, and cultural information, which are time-consuming and error-prone to interpret manually. This research investigates image processing algorithms and deep learning models for enhancing these historical documents. Using image processing techniques, we improve symbol readability and visibility, while our deep learning approach aids in reconstructing degraded content and identifying patterns. This work contributes to improving the quality of historical document analysis, particularly for graphical symbol interpretation in paleographical studies.
Purpose: Author Name Disambiguation (AND) is a critical task for digital libraries aiming to link existing authors with their respective publications. Due to the lack of persistent identifiers and intrinsic linguistic challenges such as homonymy, deep learning algorithms have become widespread for addressing this issue. This paper provides a systematic review of state-of-the-art AND techniques based on deep learning within the timeframe 2016-2024, filling a gap in existing literature. Methods: We conducted a systematic literature review using Google Scholar with keywords "author name disambiguation" + "deep learning" and temporal filters. The search yielded 52 documents, of which 28 were selected after full-text assessment. We categorized approaches based on learning paradigms: supervised, unsupervised, and hybrid methods. Additionally, we performed bibliometric analysis using OpenCitations data with network visualization via Gephi. Results:We identified 28 relevant studies employing deep learning for AND. The analysis of the comparable studies reveals that hybrid approaches achieve the best performance, with the top-performing method reaching 89.7 F1-score on AMiner datasets. Supervised methods predominantly focus on author assignment tasks, while unsupervised approaches excel at author grouping. Conclusion: Deep learning methods have significantly impacted AND by enabling integration of structured and unstructured data. Hybrid approaches effectively balance supervised and unsupervised learning, demonstrating superior performance. However, heavy reliance on AMiner datasets and lack of standardized evaluation frameworks remain critical challenges for the field’s advancement.
Every day, social media and content sharing platforms receive numerous training videos across a multitude of sectors, including home gardening, agriculture, and indoor farming. Recently, the Metaverse has offered the opportunity to create enclosed virtual spaces where a user can focus and learn specialized skills, e.g., medical, in an easy, interactive, and engaging way. However, finding suitable educational spaces for specific themes remains a challenge. In light of these advancements and to support agricultural education, in this work, we introduce AgriMus project, which aims to design agricultural-themed museums covering a broad range of topics. Users can explore and retrieve specific educational experiences through visual or textual queries. Beyond creating a dataset for the agricultural setting, we also propose a hierarchical method to model these virtual museums. Several experiments have been conducted to evaluate the effectiveness of the proposed approach. The full source code and dataset used in this study are available at: https://github.com/aliabdari/AgriMus/ .
The rapid digitization of cultural heritage collections, especially those featuring Arabic-script texts, presents distinct challenges related to access, discoverability, and long-term preservation. This paper proposes a metadata-driven framework for Arabic-script digital libraries, which is currently being explored through initial prototyping as part of the Digital Maktaba project. The framework leverages validated metadata from the Diamond catalogue and the La Pira Library’s extensive collection. To address the technical complexities of Arabic script, including calligraphy, diacritics, and ligatures, the project employs frontispiece images and the Kraken OCR engine within the eScriptorium platform to train high-accuracy recognition models. The cataloging workflow is structured around international standards such as Dublin Core and is informed by both topographic and thematic classification practices. As part of ongoing development, the project is evaluating the use of large language models (LLMs), including Arabic-specialized models, to assess their potential for extracting semantic metadata from digitized texts. This effort aims to enrich subject headings and improve classification depth, contributing to AI-assisted indexing and enhanced resource discovery. By combining human-validated metadata with machine learning pipelines, the Digital Maktaba project aims to provide a scalable, standards-aligned approach for building Arabic digital libraries, with broader applicability to other underrepresented language collections.
