Today, tourists increasingly rely on algorithmic decision-support, i.e., recommender systems, to discover points-of-interest, e.g., destinations, accommodations, or attractions, that match their preferences (Borràs et al., 2014; Ricci, 2022; Sánchez and Belloǵın, 2022). However, the influence of such algorithms extends beyond individual tourists as they can redirect tourism flows and impact local communities, businesses, and the environment (Balakrishnan and Wörndl, 2021). While these systems are designed to optimize end-user satisfaction, they struggle to ensure multistakeholder fairness due to a lack of understanding of the complex trade-offs between various stakeholders (Sonboli et al., 2022; Atzenhofer-Baumgartner et al., 2024, 2025; Burke et al., 2024, 2025). Failing to consider stakeholder needs contributes to issues such as overtourism, environmental pollution, unaffordable housing for residents, or unfair distribution of the economic benefit across stakeholders (Banerjee et al., 2025). A key limitation is the way fairness is incorporated into algorithmic decision-support systems: The computer science community primarily frames fairness through quantifiable criteria, often focusing on algorithmic approaches (Deldjoo et al., 2024) that mitigate discrimination or bias, e.g., popularity bias (Kowald et al., 2020; Müllner et al., 2023; Kowald and Lacic, 2022), which favors popular destinations over others (Rahmani et al., 2022b; Forster et al., 2025). This results in algorithmic decision-support systems that fail to implement the complex fairness criteria that are required for the numerous stakeholders in tourism. In contrast, the tourism management community tends to consider fairness as a complex, multidimensional issue that involves diverse stakeholders with competing interests. The lack of interdisciplinary collaboration results in algorithmic decision-support systems that struggle to adequately incorporate multistakeholder fairness, as they fail to account for the complex relationships between tourists, local residents, governments, and businesses. Additionally, this lack of interdisciplinary collaboration hinders a concrete operationalization of fairness goals into metrics and algorithms (Smith et al., 2023), which can be integrated into decision support and recommendation frameworks (e.g., Lacic et al., 2014; Tourani et al., 2024).

To better connect the algorithm-focused computer science research with the findings from tourism management, we review literature from two domains and investigate how they compare. Specifically, we address RQ1: How do tourism management and computer science differ in their fairness definitions?and RQ2: How can algorithm design benefit from the research body on tourism management? Through a semi-systematic literature review, we analyze existing research, highlighting differences and similarities in their fairness definitions. Our findings suggest that the tourism management community qualitatively studies multistakeholder fairness, incorporating diverse stakeholders with competing interests. In contrast, the computer science community focuses on quantitative fairness criteria that can be included in algorithmic decision-support systems, but covers only a narrow perspective on fairness. Overall, we hope that the publication at hand underscores that collaboration among the two research communities is crucially needed for developing algorithmic decision-support systems that can successfully balance the needs of numerous stakeholders in the ever-growing tourism industry.

Our work is structured as follows: Section 2 details the review methodology used to identify and filter relevant literature. Section 3 synthesizes findings from the tourism management perspective, whereas Section 4 presents the algorithmic perspective from computer science. Finally, Section 5 discusses the key differences and intersections between the two domains, and identifies open research directions.

In tourism management, effective multistakeholder governance is crucial for balancing the competing interests of various groups, such as local communities, businesses, and government entities. (Ikhtiagung and Radyanto 2020) highlight the importance of stakeholder collaboration in ecologically sensitive areas, emphasizing the need for local businesses and communities to actively participate in tourism planning to improve quality of life and community satisfaction. In this respect, (Jamal et al. 2002) investigate collaborative planning processes in protected areas, such as national parks, where stakeholders must be willing to compromise to align their diverse objectives. A similar approach is outlined by (Plummer and Fennell 2009), who emphasize the benefits of adaptive co-management in tourism, arguing that such frameworks allow stakeholders to contribute meaningfully to decision-making, even when decision-makers cannot fully comprehend all trade-offs due to bounded rationality. Moreover, (Sarhan et al. 2024) explore how sustainable tourism can be achieved through well-constructed multistakeholder partnerships, noting that such partnerships should be evaluated through a framework that includes ecological, economic, social, and cultural impacts. (Haddock-Fraser and Hampton 2012) study dive tourism in Malaysia and highlight how different stakeholders along the value chain perceive the impacts of tourism, with some focusing on economic gains while neglecting environmental sustainability. Related research also highlights the importance of responsible research, advocating for multistakeholder involvement to address societal issues beyond academic goals (Pereira-Moliner and Molina-Azorín, 2024).

Empowering local communities is essential for ensuring that tourism provides tangible benefits to the people who live in tourist areas: (Mudzengi et al. 2018) argue that although tourism can offer benefits such as employment and infrastructure improvements, local communities often struggle to capitalize on these opportunities due to limited entrepreneurial skills, lack of capital, and inadequate awareness. In contrast, (Samal and Dash 2025) show how ecotourism in India has enhanced local livelihoods, with a focus on multistakeholder participation and the development of infrastructure and capacity-building programs. Furthermore, research explores how small and medium-sized enterprises can diversify local economies and reduce reliance on tourism through digital innovation and government support (Hasayotin et al., 2024). The work by (Sitikarn et al. 2022) also sheds light on the value of community-based tourism, using the example of coffee production in Thailand, where local involvement and innovation have led to the creation of diverse tourism offerings that benefit the community both economically and socially. (Chan 2023) discuss how sustainable practices in rural tourism are driven by local community collaboration, government support, and a focus on environmental and social development. Similarly, Yudistira et al. (2023) use user studies and predictive modeling to estimate regional development, suggesting that local involvement and long-term benefits should be prioritized. Research also emphasizes the role of community engagement in tourism development, noting that mapping community assets can help identify resources that support sustainable social, economic, and environmental development (Yeager et al., 2024). (Romeo et al. 2024) present an initiative, which emphasizes the critical role of mountain regions for sustainable tourism development and supports value chain development to ensure that benefits flow equitably to all regional stakeholders. (Sigala 2021) reviews literature on sharing and platform economy in tourism, emphasizing the roles of key stakeholders: platforms, providers, users, and the environment. Digital sharing platforms like Airbnb and Uber are rich financial opportunities for locals, but can lead to complex socioeconomic, or ethical issues, e.g., gentrification. Finally, (Adams and Infield 2003) explore the challenges associated with revenue distribution in tourism, highlighting the competing interests of local, national, and international stakeholders, and pointing out the difficulties in ensuring that local communities receive fair compensation for their participation in tourism.

Sustainable tourism development seeks to balance the needs of tourists with the preservation of the environment as outlined by (Higgins-Desbiolles 2018). The authors argue that the tourism industry's current growth-oriented mindset is incompatible with true sustainability, suggesting that tourism must respect ecological and social limits. The research by (Trang et al. 2023) explores how ecotourism can be promoted effectively in Vietnam, emphasizing the role of so-called Destination Management Organizations in managing local challenges and improving tourist satisfaction. Moreover, (Majdak and de Almeida 2022) propose distributing tourism across rural areas to reduce overtourism in popular destinations, arguing that this strategy brings economic benefits to less-visited areas and leads to a fairer distribution of tourists. Research also shows that when tourists believe that all stakeholders are treated in a fair way, they are more likely to engage in prosocial behaviors, such as positive word-of-mouth, which can further support sustainable tourism practices (Rodriguez-Sanchez et al., 2025). In the domain of thermal tourism, (Blanco-Cerradelo et al. 2022) identify several factors that impact the sustainability of tourism across economic, environmental, and social dimensions, underscoring the need for a holistic approach to tourism management. (Su 2020) highlight that there is a trade-off between protecting cultural sites and exploiting touristic value. Plus, stakeholders, such as heritage conservation groups, tourist agencies, or local business and residents, differ in their objectives and values, which makes it hard to find a common ground in decision-making processes. Finally, (Khatri and Sharma 2024) use questionnaires to investigate barriers and challenges of tourism stakeholders, such as hoteliers or tour operators. The key challenges include the protection of the environment and cultural assets, and the incorporation of uniform socio-cultural, and techno-environmental constraints across all stakeholders.

(Biswas et al. 2021) emphasize that most recommender systems are optimized for user satisfaction, which can result in unfair exposure for items, such as points-of-interest (POIs). Conversely, systems optimized for item exposure may lead to unfair user experiences. To address this two-sided fairness problem, they propose an algorithm that guarantees a minimum exposure level for items while distributing the loss in recommendation quality evenly across users, which ensures envy-freeness (Arnsperger, 1994; Burke et al., 2022). Similarly, (Wu et al. 2020) tackle fairness under capacity constraints, proposing a reranking strategy that modifies recommendation lists over multiple rounds. This ensures fairness while accounting for the limited capacity of venues like restaurants. (Rahmani et al. 2022b) show that POI recommender systems often suffer from popularity bias, favoring active users and popular destinations. While many models perform well in accuracy, they fail to provide balanced fairness across users and items. A related bias is temporal: (Rahmani et al. 2022c) demonstrate that users seeking leisure-time recommendations receive preferential treatment over those searching during work hours, despite equal interaction histories. Moreover, research identifies position bias in POI recommender systems, where nearby but lower-quality venues are ranked higher than more distant yet highly rated establishments, negatively impacting the exposure of deserving businesses (Banerjee et al., 2020). Additionally, (Merinov and Ricci 2024) simulate tourists' limited knowledge of tourist sites and demonstrate that personalization can be balanced with sustainability to some degree. They show that a standard recommender system can promote less-visited locations while maintaining user satisfaction. In the context of personalized, scenario-specific travel recommendations, related research introduces SAR-Net (Shen et al., 2021). This model uses attention mechanisms and scenario-specific transformations to improve fairness and accuracy, while its fairness coefficient corrects exposure bias across different user scenarios. (Rahmani et al. 2022a) also tackle fairness and overtourism with a recommender system optimized for item and user fairness. They employ metrics such as Coverage (Silveira et al., 2019), Novelty (Gunawardana et al., 2012), Generalized Cross-Entropy (Deldjoo et al., 2021), and Mean Absolute Deviation (Melchiorre et al., 2021) to evaluate fairness and identify popularity and bias. (Balakrishnan and Wörndl 2021) examine how tourism recommender systems can incorporate the needs of tourists, locals, and service providers. They discuss inter- and intra-stakeholder dynamics and show, through user studies, that users are sensitive to other stakeholders' needs. External influences such as legislation or seasonal effects are also considered. (Banerjee et al. 2023) argue for balancing the sometimes competing interests of different stakeholders, highlighting the importance of fairness not only for users and providers but also for the broader society and environment. (Patro 2023) examine fairness in recommendation systems on online platforms that include multiple stakeholders. Mainly, they find that most fairness research has focused on settings with two stakeholders, and that improving individual fairness for one stakeholder group often reduces utility for another stakeholder group. This highlights the need for further exploration of multistakeholder fairness involving three or more groups, where the trade-offs become more complex and less well understood.

Several studies aim to align tourism recommendations with sustainability goals. (Banerjee et al. 2025) propose a recommender system that incorporates CO₂ emissions, destination popularity, and seasonality into travel recommendations. Their user study confirms that users are willing to trade utility for sustainability. (Merinov et al. 2022) design itineraries that avoid overcrowded POIs while preserving user satisfaction by estimating both utility and environmental impact. Similarly, recent research presents a recommender system that promotes societal fairness by recommending environmentally friendly and seasonally balanced destinations (Banerjee et al., 2024). (Banik et al. 2023) explore user perceptions of sustainability in Venice and find that offering one sustainable alternative per unsustainable choice increases user satisfaction, especially when accompanied by explanations. Furthermore, (Banerjee 2023) provide a roadmap for incorporating societal fairness into tourism recommender systems by balancing stakeholder concerns, including those of local residents and the environment. Related research also explores how to reduce popularity bias and crowding through time-sensitive, stakeholder-aware recommendation techniques (Merinov, 2023). A broad overview is offered by (Banerjee et al. 2023), who review fairness in tourism recommender systems from the perspectives of users, providers, items, and society, identifying a gap in fairness research pertaining to issues such as society and sustainability.

Data sparsity is especially prevalent for novel, emerging tourist destinations and therefore, recommender systems often fail to recommend such destinations. To resolve this issue, (Solano-Barliza et al. 2024) develop a rule-based system that uses hierarchical criteria like distance and cuisine to recommend restaurants, especially in emerging tourist regions where data is limited. Similarly, (Sánchez and Bellogín 2021) suggest merging datasets from different cities and distinguishing between tourists and locals to improve POI recommendations. However, their findings indicate that tourists often benefit more from such systems than locals, which can be perceived as unfair. (Khaili et al. 2024) propose a multi-funnel architecture to address the cold-start issue (Wei et al., 2021; Lacic et al., 2015) for newly listed items in travel platforms. By separating and then merging cold-item rankings with regular items, their system improves platform diversity and long-term partner retention with minimal immediate performance losses.