[Paper Review] Value Identification in Multistakeholder Recommender Systems for Humanities and Historical Research: The Case of the Digital Archive Monasterium.net

Florian Atzenhofer-Baumgartner, Bernhard C. Geiger+2 more|arXiv (Cornell University)|Sep 26, 2024

Recommender Systems and Techniques4 citations

TL;DR

This paper identifies and analyzes the diverse values of multiple stakeholders—such as researchers, editors, publishers, and funding agencies—in the context of recommender systems for digital humanities archives, using Monasterium.net as a case study. It proposes a value-aware framework to guide the design of multistakeholder recommender systems that balance conflicting objectives, enhance scholarly discovery, and support sustainable cultural heritage curation.

ABSTRACT

Recommender systems remain underutilized in humanities and historical research, despite their potential to enhance the discovery of cultural records. This paper offers an initial value identification of the multiple stakeholders that might be impacted by recommendations in Monasterium.net, a digital archive for historical legal documents. Specifically, we discuss the diverse values and objectives of its stakeholders, such as editors, aggregators, platform owners, researchers, publishers, and funding agencies. These in-depth insights into the potentially conflicting values of stakeholder groups allow designing and adapting recommender systems to enhance their usefulness for humanities and historical research. Additionally, our findings will support deeper engagement with additional stakeholders to refine value models and evaluation metrics for recommender systems in the given domains. Our conclusions are embedded in and applicable to other digital archives and a broader cultural heritage context.

Motivation & Objective

• To address the underutilization of recommender systems in digital humanities and cultural heritage research, particularly in the context of historical document discovery.
• To identify and analyze the diverse, often conflicting, values and objectives of multiple stakeholders involved in digital archives like Monasterium.net.
• To lay the groundwork for designing value-aware recommender systems that support scholarly access, curation, and long-term sustainability in non-commercial cultural heritage contexts.
• To extend the discourse on recommender systems beyond museum-centric applications to research-oriented digital archives with complex stakeholder ecosystems.
• To support the development of more nuanced evaluation metrics and stakeholder-informed system design in digital humanities.

Proposed method

• Conducted a stakeholder analysis focusing on six key stakeholder groups: researchers, editors, aggregators, platform owners, publishers, and funding agencies.
• Mapped stakeholder objectives and values using qualitative insights from domain experts deeply involved in Monasterium.net and related digital humanities initiatives.
• Identified value conflicts such as those between scholarly integrity (researchers) and platform growth (platform owners), or between public engagement (publishers) and archival accuracy (editors).
• Proposed a multistakeholder value-aware framework for recommender system design, emphasizing fairness, transparency, and sustainability in non-commercial settings.
• Used insights from the ERC-funded 'From Digital to Distant Diplomatics' project to ground the analysis in an ongoing, real-world system redevelopment.
• Laid the foundation for future empirical validation through structured interviews and user studies with broader stakeholder groups.

Figure 1: Example document: Recto side of a charter by Holy Roman Emperor Frederick II to Babenberg Duke Frederick II containing the text of the Privilegium maius - itself a forged adaptation of the Privilegium minus by Frederick I -, confirming it, and extending its terms. Purportedly issued in Ver

Experimental results

Research questions

• RQ1 What are the key values and objectives of diverse stakeholders in digital humanities archives like Monasterium.net?
• RQ2 How do conflicting stakeholder interests—such as scholarly accuracy versus platform visibility—affect the design of recommender systems in cultural heritage contexts?
• RQ3 In what ways can recommender systems be adapted to serve both expert researchers and the general public in non-commercial, knowledge-focused domains?
• RQ4 What role do funding agencies and policymakers play in shaping the ethical and sustainable deployment of recommender systems in digital archives?
• RQ5 How can value-aware design principles be operationalized to guide the development of evaluation metrics and system features in multistakeholder settings?

Key findings

• The study identified distinct and often conflicting values among stakeholders: researchers prioritize scholarly accuracy and discovery, while platform owners focus on growth and user engagement.
• Publishers and media stakeholders value public engagement and narrative storytelling, while funding agencies emphasize accessibility, sustainability, and academic reuse of digitized collections.
• Policymakers and cultural heritage institutions are concerned with legal, licensing, and ethical constraints that affect recommendation fairness and long-term access.
• Current full-text search mechanisms on Monasterium.net are insufficient for high-dimensional, complex archival data, highlighting the need for intelligent recommendation systems.
• The integration of recommender systems in digital humanities archives requires a shift from commercial, revenue-driven models to value-sensitive, stakeholder-informed design approaches.
• The paper establishes a foundational framework for future research, including stakeholder-driven metrics and evaluation methods, essential for responsible AI deployment in cultural heritage.