Abstract
Introduction
In the rapidly advancing field of urban mobility, shared micromobility systems, such as dockless bicycles and electric scooters, have emerged as innovative solutions. Offering rentable, accessible, and flexible transportation options, these systems have reshaped urban mobility in densely populated cities (Fishman et al., 2015; Gössling 2020; S. A. Shaheen et al., 2010). Shared micromobility systems offer various advantages, including integration with public transportation, promoting modal shifts, and providing alternatives for short distances (Fishman & Cherry, 2016; Öztaş Karlı & Çelikyay, 2022; S. A. Shaheen et al., 2010).
Shared micromobility refers to transportation services that provide users with short-term access to lightweight vehicles, such as bicycles, e-bikes, and e-scooters. Users typically access these services through digital platforms, allowing them to pick up and drop off vehicles at various locations without fixed stations. Their design helps address urban issues, such as congestion and first- and last-mile connectivity challenges.
The popularity of these systems highlights their current significance and suggests a surge in research and debate in this area. The literature on shared micromobility spans disciplines from urban planning and engineering to social sciences, law, environmental science, and health (Cicchino et al., 2021; Félix et al., 2020; Hardt & Bogenberger, 2019; Lo et al., 2020; Reck et al., 2020). This diversity reveals the complex and multifaceted nature of shared micromobility within urban discourse. Given this context, the critical role of literature reviews in advancing understanding of this rapidly evolving field is evident.
Despite this growing body of literature, no study to date has comprehensively synthesized the thematic evolution of shared micromobility using scalable, automated, and theory-neutral approaches. This gap justifies the need for a structured, data-driven review of the field.
In this context, the study provides the first large-scale, data-driven synthesis of the shared micromobility literature, revealing how academic attention has evolved alongside real-world adoption and policy transformations. By mapping the thematic structure and evolution of this field, the study offers a deeper understanding of how shared micromobility supports sustainable, inclusive, and efficient urban transport systems. These insights are valuable to scholars as well as to urban planners and policymakers seeking to integrate shared micromobility into broader mobility transitions.
The emergence and implementation of shared micromobility systems gained momentum after 2010, driven by advancements in GPS tracking, smartphone integration, and digital payment systems. This period marks the global shift from experimental pilots to commercial-scale services in urban centers. Consequently, the period from 2010 to 2023 reflects a critical phase of diffusion, adoption, regulation, and scholarly attention in the field.
Latent Dirichlet Allocation (LDA) topic modeling is a sophisticated data analytics technique, proven effective in synthesizing and interpreting academic works to identify common research trajectories and intellectual contours. LDA is a generative probabilistic model used to discover abstract topics within a collection of documents. It assumes that documents are composed of multiple topics, with each topic represented by a distribution over words. This allows researchers to uncover latent semantic structures in large-scale textual corpora (Blei et al., 2003). This method excels at distilling the essence of research to illuminate patterns and thematic concentrations within bodies of literature (Chappelle et al., 2024; Rodrigues et al., 2017; Taghandiki & Mohammadi, 2023).
The application of LDA in this context is appropriate because it enables an objective mapping of the field without relying on prior assumptions or manual coding. Given the multidisciplinary and rapidly expanding shared micromobility research, LDA offers a scalable and replicable way to detect prevailing and emerging themes that would otherwise remain hidden.
This study, therefore, critically assesses the thematic focus of shared micromobility literature by evaluating research intensity across topics and identifying fundamental research areas. Additionally, this work seeks not only to unveil dominant research topics but also to track their evolution, providing a structured and in-depth examination of the field. Through this methodological approach, the article contributes a comprehensive thematic map of the shared micromobility literature, thereby enriching the understanding of past research trajectories and informing future scholarly endeavors.
Accordingly, this research addresses the following key questions:
These questions align with the study’s goal of unveiling the intellectual contours of shared micromobility research systematically and data-drivenly. Together, they form an integrated framework for capturing bibliometric structure (RQ1), thematic content (RQ2), and temporal change (RQ3).
To set the stage for an in-depth examination of shared micromobility, the following sections provide a comprehensive analysis of the topic. Following this introduction, the article systematically outlines the methodology, then presents a detailed account of the thematic findings. Subsequently, the discussion intertwines these findings with broader implications for shared micromobility research, and the conclusion reflects on the study’s insights, limitations, and future research.
Methods
This research is based on bibliometric analysis and topic modeling methodologies. Bibliometric analysis summarizes and quantifies the information contained in scientific publications. It provides quantitative insights into various aspects of academic output, including leading researchers, geographical distribution of authors, prominent institutional affiliations, chronological research trends, and citation growth (Gireesh & Gowda, 2008). By quantitatively assessing these dimensions, bibliometric analyses recognize prevailing and emerging trends within a given research field (Ayaz et al., 2021).
Topic modeling, on the other hand, is a machine learning technique used to uncover hidden semantic structures within large textual datasets (Ayaz et al., 2023; Blei et al., 2003; Özköse et al., 2023). It assumes that documents consist of latent topics, each represented by a probabilistic distribution of words. Among the existing models, Latent Dirichlet Allocation (LDA) is favored for its robustness, interpretability, and ability to reveal the evolving intellectual structure of a research domain (Blei, 2012; Blei et al., 2003).
In this study, LDA was employed to extract latent thematic clusters from the corpus and to analyze their evolution. By identifying probabilistic co-occurrence patterns of terms, LDA enables the detection of dominant and emerging themes, and shifts in academic focus, thereby offering a comprehensive thematic mapping of the shared micromobility literature.
While several alternative algorithms, such as Hierarchical Latent Dirichlet Allocation (HLDA), Hierarchical Dirichlet Process (HDP), Non-Negative Matrix Factorization (NMF), and Correlated Topic Model (CTM), exist, they often require complex parameter tuning and may yield inconsistent topic coherence (Gurcan & Cagiltay 2022; Vayansky & Kumar 2020). In contrast, LDA offers greater flexibility in determining the optimal topic count through iterative testing, ensuring semantic stability and analytical reliability. Because of its methodological rigor and empirical robustness, LDA remains one of the most widely adopted approaches in text mining and knowledge mapping research (Blei, 2012).
To ensure academic rigor, the LDA implementation in this study followed a structured multi-stage process: corpus construction, text pre-processing, iterative parameter tuning, model validation, and thematic interpretation. This process aligns with established best practices in topic modeling literature.
Selection Criterias, Search Strategies and Data Collection
The selection of a bibliometric database is a crucial first step for precision and relevance in shaping a research dataset. In this study, an initial investigation of the prevalence of shared micromobility literature in two major bibliometric databases (Web of Science [WoS] and Scopus) was conducted. These platforms are widely recognized for their comprehensive coverage of scientific articles. The analysis revealed that Scopus included a greater number of articles than WoS. This discrepancy stems from the broader journal coverage of Scopus, which covers almost all journals indexed by WoS (Mongeon & Paul-Hus, 2016). Based on this finding, Scopus was chosen as the preferred bibliometric database for this study, and the search strategy was developed accordingly.
To comprehensively investigate the topic of “shared micromobility” within the Scopus database, key search terms were identified based on an analysis of studies cited in the existing literature (Abduljabbar et al., 2021; Teusch et al., 2023; Zhu et al., 2022). These terms were used to construct a query targeting the titles, abstracts, and keywords of documents indexed in Scopus. The search was refined to include only peer-reviewed journal articles (research and review types) to ensure relevance and scientific rigor. The temporal scope covered English-language publications from the beginning of the available records through the end of 2023, with articles from 2024 deliberately excluded to maintain a consolidated and analyzable body of knowledge. While this decision ensures linguistic consistency and access to peer-reviewed international sources, it also is a methodological limitation, as relevant studies published in other languages (e.g., French, Spanish, or Chinese) may have been omitted. This range was selected because it represents the period during which shared micromobility evolved from early pilot projects to mature, large-scale systems, and academic interest expanded accordingly. Covering this full diffusion phase ensures a comprehensive understanding of the field’s intellectual evolution.
Based on these parameters, we formulated the following query to retrieve the most relevant articles on shared micromobility from the Scopus database:
“shared micromobility” OR “shared e-scooter” OR “sharing e-scooter” OR “e-scooter sharing” OR “shared bike” OR “shared e-bike” OR “e-scooter sharing” OR “shared bicycle” OR “shared e-bicycle” OR “sharing e-bicycle” OR “bicycle sharing” OR “dockless bike sharing” OR “dockless bike share” OR “dockless bike-sharing” OR “bike-sharing” OR “sharing bike” OR “shared electric bike” OR “e-moped sharing” OR “shared e-moped”
This query was executed on March 13, 2024. A total of 2,021 articles were retrieved for the period from 2010 to 2023: 1,984 research articles and 37 review articles. The title, abstract, and author keyword information from these articles were compiled into a dataset, establishing a comprehensive body of literature on shared micromobility.
Pre-Processing of Data, Application of the Topic Model and Data Analysis
Topic modeling, an integral part of text mining, requires a series of preparatory steps to transform the raw data into a format suitable for analysis (Aggarwal & Zhai, 2012). The preliminary steps are summarized below:
Initially, a corpus of stop words is created and used to filter the dataset. Stop words (commonly occurring terms such as “a,”“an,”“is,”“the,”“of,” and “for”) are removed because they do not contribute meaning to the analysis. Following the elimination of stop words, the corpus undergoes a cleaning and tokenization phase. During this stage, all text is normalized to lowercase to ensure consistency, and embedded special characters and punctuation are removed.
Next, the preprocessing stage includes lemmatization, which aims to separate words into their root forms by eliminating affixes, thereby standardizing the dataset for analysis. Once the pre-processing steps (comprising stop word removal, text normalization and cleaning, and lemmatization) are completed, the titles, abstracts, and keywords of the documents are merged to construct the corpus for topic modeling analysis.
A preliminary LDA analysis on this corpus revealed that the terms “shared micromobility” and related query phrases appeared too frequently across topics. Given their direct relevance and limited discriminatory power, these terms were added to the stop word list to prevent distortion in topic representation.
The LDA model was then iteratively refined through experimental runs. Several models were generated by varying the number of topics (
Topic modeling analysis was performed using the ldaMulticore model in the Python Gensim library (Prabhakaran, 2018). To fine-tune the LDA-based topic model, parameter values were carefully selected to optimize performance. Initial settings included both symmetric and asymmetric configurations for α (topic distribution within documents), a symmetric setting for β (word distribution within topics), and hyperparameters workers = 13, random_state = 42, and pass = 20. Consistent with prior literature, we followed an iterative and heuristic calibration process to ensure optimal results (Mimno et al., 2011).
PyLDAvis (a visualization tool for topic modeling) was employed to identify thematic clusters within the dataset (Mabey, 2023; Prabhakaran, 2018). The lambda (λ) value (which indicates the significance of words within each topic) was set at 0.6 in accordance with established standards (Ayaz et al., 2023; Sievert & Shirley, 2014). Through a collaborative assessment, we confirmed the semantic coherence of the 12 identified topics and assigned descriptive labels based on the dominant keywords.
Subsequent analytical steps involved calculating the proportion of each topic per document, analyzing the distribution of terms within topics, and examining topic prevalence across the corpus. These distributions were used for the analysis of temporal trends through slope calculations. This comprehensive method produced a catalog of the top 30 terms (ranked by frequency) representative of each topic.
Findings
The exploration of shared micromobility literature through LDA reveals the topics and trends that have shaped the field from 2010 to 2023. This section presents the quantitative and thematic findings of our analysis, offering a comprehensive view of the evolution and current state of shared micromobility research. By presenting data on publication trends, key contributors, thematic distributions, and the interplay between research focuses, we provide a detailed snapshot of the field’s development.
These findings reflect not only the descriptive scope of the dataset but also the thematic structures identified through probabilistic modeling. By linking bibliometric indicators with LDA-based topic modeling, we derive both empirical patterns and conceptual insights that collectively map the intellectual contours of shared micromobility research.
Figure 1 depicts the trend of article numbers on shared micromobility by year. According to the graph, the number of publications (starting with only four articles in 2010) increased steadily over the years. The increase between 2016 and 2021 is particularly noticeable. After peaking in 2021 with 381 articles (19% of the total), the number slightly declined by about 1% to 358 articles in 2023.
Number of articles by year.
This growth trajectory underscores the transformation of shared micromobility from a niche innovation into a prominent academic research domain. The surge in publications corresponds to the rapid global expansion of micromobility services and growing policy interest. The recent decline may indicate a phase of thematic stabilization or a shift in scholarly focus toward new areas.
According to Figure 2, the most prolific author is Ma X. (Tianjin, China), with 15 articles. Other significant contributions have been made by Szeto W.Y. (Shenzhen, China), with 14 articles, and Axhausen K.W. (Zurich, Switzerland), with 13 articles. Several other authors also made notable contributions, including Ji Y., Chen X., Cheng L., and Chen J. (Nanjing, China); Li A. (Beijing, China); Yang L. (Chengdu, China); and Eluru N. (Orlando, United States).
The top contributing authors.
The prominence of these authors reflects the presence of strong research clusters, particularly in East Asia. Their sustained output suggests the existence of institutional and regional centers of excellence that are actively shaping the intellectual contours of shared micromobility research.
Social Sciences holds the largest share with 21.1% and 1,047 articles, making it the most represented topic area. Engineering follows with 19.5% and 968 articles. Environmental Science and Computer Science both account for 12% of the share. The least represented include more specialized disciplines or those less directly related to shared micromobility, such as Agricultural and Biological Sciences, Arts and Humanities, Chemistry, and Neuroscience (Table 1).
Topic Areas of the Shared Micromobility Articles.
The strong presence of Social Sciences highlights the human-centered dimension of shared micromobility, particularly in areas such as user behavior, policy, and societal impact. Meanwhile, the significant contributions from Engineering and Environmental Science reflect a growing interest in technological innovation and sustainability. The comparatively low representation from the natural sciences suggests underexplored interdisciplinary opportunities for future research.
Sustainability (Switzerland) emerges as the leading journal in this field, accounting for 11.3% of the articles.
Publication Sources.
The National Natural Science Foundation of China leads the field with 462 publications, holding a 27.2% share. The Fundamental Research Funds for the Central Universities and the National Key Research and Development Program of China follow, contributing 4.6% and 4.5%, respectively. The National Science Foundation in the USA ranks fourth with a 2.9% contribution (Table 3). The concentration of funding and research output in China underscores the country’s key role in shaping the global trajectory of shared micromobility research. This pattern reflects strategic national investments and policy priorities aimed at advancing sustainable urban transportation.
The Top Contributing Institutions.
According to Table 4, China leads the field with 830 publications, accounting for 29.6% of the total, which makes it the leader. The United States follows with 414 publications (14.8%), and the United Kingdom ranks third with 153 publications (5.5%). Other countries contribute 2.5% to 3.3% to this field. This distribution highlights regional disparities in research activity, shaped by differing policy environments, urban transport priorities, and the extent of micromobility deployment.
Top Contributing Countries.
As seen in Table 5, the three most researched topics are “Travel Patterns,”“Technology Acceptance,” and “Innovative Mobility Services” respectively. The three least studied topics are “Trip Choice,”“Demand Prediction,” and “Safety” in that order. These thematic clusters were derived through iterative Latent Dirichlet Allocation (LDA) analysis, based on the probabilistic co-occurrence of terms. They were subsequently labeled through author consensus by examining the dominant keywords in each topic.
The 12 Topics Discovered By LDA.
While the top themes reflect strong scholarly interest in user behavior and innovation adoption, less-covered themes, such as “Safety” and “Demand Prediction” remain critical for real-world system design and policymaking. Their scarcity in the literature highlights a notable thematic gap and underscores the need for deeper investigation into operational risk management, predictive analytics, and infrastructure safety.
Table 6 reveals notable shifts in topic interest over time. The topic of “Travel Patterns” has exhibited a consistent upward trend, emerging as the most researched area with 484 publications. This underscores that travel habits and mobility behaviors constitute a major focus within the shared micromobility literature. The “Technology Acceptance” topic (which initially attracted limited scholarly attention) experienced a sharp increase in research activity, particularly during 2020 and 2021. The topic of “Innovative Mobility Services” has also shown a steady rise in interest, reflecting a focus on the development of new micromobility solutions and service models. Topics such as “Decision-Making Processes,”“Transportation Policies,” and “System Management” have demonstrated gradual growth, peaking in 2021. “Travel Modes and Transportation” has seen a regular increase, becoming a prominent research domain in recent years. Topics related to Regulation and Safety have likewise gained momentum over time, with “Safety” in particular receiving heightened attention in recent studies. Technical areas like “System Optimization” and “Fleet Management and Rebalancing” also peaked in 2021. Meanwhile, topics such as “Environmental Impact,”“Demand Prediction,” and “Trip Choice” remain relatively underexplored.
Distribution of the Number of Publications on the Topic According to Years.
The longitudinal distribution of research themes suggests that developments (such as the COVID-19 pandemic, smart city initiatives, and the rise of electric vehicle infrastructure) have significantly shaped scholarly interest. In particular, the post-2020 surge in “Technology Acceptance” appears to correlate with the widespread adoption of digital mobility platforms during and after the pandemic.
The “Travel Patterns” topic holds the largest share (21.92% of the volume) and has shown significant growth with a momentum rate of 8.54%, indicating that it is not only a dominant area but also a rapidly growing focus of scholarly interest (Figure 3).
Ranking by volume and accelerations.
In this context, “volume” denotes the proportion of total documents assigned to a topic, reflecting its cumulative presence in the literature, while “momentum” refers to the rate of increase in publications over time, signaling emerging attention. With a volume of 17.86% and momentum of 5.70%, the “Technology Acceptance” topic emerges as the second-fastest-growing area in the field. This reflects the academic community’s increasing interest in technological innovation adoption and its influence on user behavior. The “Innovative Mobility Services” topic ranks third in both volume (13.81%) and momentum (3.35%), highlighting concentrated research on service design and platform-based mobility innovations. The “Decision Making” topic ranks fourth in volume (11.13%) but has a comparatively lower momentum rate (2.99%), suggesting a well-established body of literature with slower recent growth.
This dual-axis analysis of volume and momentum offers a nuanced understanding of thematic dynamics in the field, enabling researchers and policymakers to distinguish between mature areas of inquiry and rapidly emerging ones.
Figure 4 shows that the topics with the highest slopes, in order, are “Travel Patterns,”“Technology Acceptance,”“Innovative Mobility Services,” and “Travel Modes and Transportation.” Conversely, “Demand Prediction” and “Safety” exhibit the lowest slopes. In this context, “slope” refers to the average annual growth in topic-related publications, indicating sustained, long-term interest in a particular research area. It complements momentum by capturing persistent trends beyond short-term fluctuations.
The slope of the topics within themselves.
These gradients highlight the intensifying focus and scholarly engagement within the shared micromobility research community, helping to forecast which topics are likely to attract increased academic and policy attention. Notably, topics with steeper gradients can be viewed as strategic priorities for future research, policy formulation, and industrial innovation.
As shown in Figure 5, the negative slopes for “Decision Making” (−7.34) and “System Optimization” (−7.32) suggest that these areas are beginning to attract less scholarly interest relative to other topics, possibly indicating thematic saturation or a shift toward emerging research areas. Although “Regulations” (−3.25) remains a significant area, its relative rate of growth has declined compared to other domains. A decreasing slope does not indicate irrelevance but may reflect conceptual consolidation, thematic maturity, or absorption into broader interdisciplinary frameworks. Monitoring such patterns is essential for anticipating future thematic transitions in the field.
The order of the slopes of the topics compared to other topics.
“Trip Choice” (−0.57) and “Fleet Management and Rebalancing” (−0.42) exhibit modest negative slopes, suggesting slower recent growth. “Safety” (0.08) and “Demand Prediction” (0.50) show nearly stable levels of interest or slight increases, while “Environmental Impact” (0.83) displays a positive trajectory, emphasizing the rising importance of sustainability.
“Travel Patterns” (2.38) reflects a notable increase in attention, signifying its continued academic relevance and popularity. “Innovative Mobility Services” (3.26) and “Technology Acceptance” (4.40) demonstrate substantial momentum, highlighting the growing emphasis on digital innovation and user-centric service design. “Travel Modes and Transportation” (7.43) (with the highest positive slope) emerges as the most rapidly expanding area, indicating strong academic interest in the integration of micromobility within multimodal urban transport systems, a trend likely to shape the future research agenda.
Discussions
Shared micromobility has emerged as a dynamic research domain with the potential to redefine urban mobility paradigms through user-centered, technology-driven, and sustainability-oriented transportation models. The bibliometric and LDA-based analyses highlight diverse research trajectories, thematic concentrations, and intellectual shifts, particularly emphasizing the sharp rise in academic interest after 2016. This growing attention signals a transition toward urban mobility systems that are not only efficient and integrated but also responsive to behavioral, regulatory, and environmental considerations. These evolving patterns underscore the need for interdisciplinary frameworks to deepen the understanding of shared micromobility and to support its integration into future urban transport systems.
Bibliometric Characteristics of Shared Micromobility Research (RQ1)
Between 2010 and 2023, academic output related to shared micromobility has shown a steady increase, with particularly notable growth from 2016 to 2021. The peak in 2021 marks a significant milestone in the field’s development, potentially driven by widespread urban digitalization and the expansion of smart mobility infrastructures. The slight decline in publication volume after 2021 may be attributed to post-pandemic shifts in funding priorities or suggest a maturing phase within the field.
The prominence of authors such as Ma X. and Szeto W.Y. indicates the emergence of influential scholars who have shaped the discourse and advanced methodological approaches in this domain. Their recurring presence reflects the existence of intellectual hubs and scholarly leadership, contributing to thematic coherence and research continuity.
As shown in Table 1, the disciplinary distribution reveals that shared micromobility research is highly interdisciplinary. Social Sciences and Engineering are the most represented domains, followed by Environmental Science, Computer Science, Energy, and Decision Sciences. This distribution confirms that the field addresses not only technological challenges but also societal concerns such as user behavior, governance, and environmental sustainability.
This disciplinary evolution suggests a temporal shift in research priorities. From 2010 to 2015, early contributions focused primarily on operational and technological feasibility. During 2016 to 2019, attention broadened to include user behavior, urban planning, and service integration. The post-2020 period has seen a notable diversification of themes, including rising interest in sustainability, safety, regulations, and equity-oriented service design. This progression indicates that shared micromobility research has evolved from a niche innovation to a structured agenda within the policy and planning framework.
An examination of journals shows that Sustainability (Switzerland),
At the geographic and institutional levels, China leads the field in publication output, supported by strong institutional backing and funding. As noted in Tables 3 and 4, entities such as the National Natural Science Foundation of China and the Fundamental Research Funds for the Central Universities have invested substantially in this area. This prioritization reflects China’s national strategy for smart urbanization, low-carbon mobility, and technological leadership. In contrast, the U.S. contribution (while significant) is shaped by a more decentralized and innovation-driven model, often grounded in localized pilot projects or industry-led initiatives.
This divergence between centralized and decentralized research agendas opens up avenues for comparative studies. Future research could explore how institutional configurations, regulatory frameworks, and urban governance models influence the development and outcomes of shared micromobility systems across national contexts. Overall, the bibliometric findings highlight the evolution of shared micromobility as an academic field, moving from isolated, case-based investigations to a global, interdisciplinary, and policy-relevant domain.
Dominant Thematic Clusters in Shared Micromobility Research (RQ2)
The growing interdisciplinary scope of shared micromobility research reflects its complex position at the intersection of technology, user behavior, urban policy, and environmental sustainability. Bibliometric findings reveal a strong representation of Social Sciences, underscoring the increasing scholarly emphasis on understanding the attitudes, perceptions, and adoption behaviors of users. This observation is reinforced by the LDA topic modeling, where “Technology Acceptance” and “Travel Patterns” emerged as the dominant thematic clusters in terms of both volume and momentum.
The “Technology Acceptance” theme, supported by studies such as Akbari et al. (2020), Beale et al. (2023), and Öztaş Karlı et al. (2022), highlights the crucial role of perceived ease of use, accessibility, and affordability in the adoption of shared micromobility systems. These findings suggest the necessity of user-centered design and inclusive policy approaches to support the diffusion of these systems. Similarly, the “Travel Patterns” theme emphasizes how shared micromobility systems are reshaping short-distance travel behaviors, particularly in dense urban environments (Dibaj et al., 2021; Liao & Correia, 2022; X. Ma et al., 2020; Schwinger et al., 2022; Şengül & Mostofi, 2021).
The prominence of “Innovative Mobility Services” indicates the field’s responsiveness to user demand and evolving mobility expectations (Bieliński & Ważna, 2020; Lazarus et al., 2020). This theme captures business model innovations, app-based service designs, and responsiveness to real-time feedback, highlighting shared micromobility as a dynamic and adaptive system.
Research on “Travel Modes and Transportation” further illustrates how shared micromobility services are integrated into existing transport networks, providing first- and last-mile solutions and supporting multimodal travel (Gössling, 2020; McQueen et al., 2021; Oeschger et al., 2020; Pazzini et al., 2022; Şengül & Mostofi, 2021; Sun & Ertz, 2022; Tian et al., 2022). These developments underline the importance of infrastructure planning and transit-oriented development (TOD) policies for effectively embedding shared micromobility within broader urban transport ecosystems.
In parallel, the theme of “Environmental Impact” reflects the growing attention to the sustainability contributions of shared micromobility systems, particularly in terms of reducing greenhouse gas emissions and promoting cleaner transport alternatives (Asensio et al., 2022; McQueen et al., 2021; Şengül & Mostofi 2021). These findings emphasize the need for policy integration and investment in environmentally friendly infrastructure to amplify environmental benefits.
Operationally, “System Optimization” and “Fleet Management and Rebalancing” focus on improving service efficiency and reliability through algorithmic allocation, vehicle distribution strategies, and maintenance planning (Nikitas, 2019; Yi & Smart, 2021). These topics remain essential for ensuring user satisfaction and financial viability, especially as services scale across urban areas.
“Regulations and Safety” form another critical thematic cluster, addressing the ethical and legal dimensions of shared micromobility. Key studies (Bakker, 2018; Banister, 2008; Beale et al., 2023; James et al., 2019; Q. Ma et al., 2021; Schnieder et al., 2021; S. Shaheen & Cohen, 2019) emphasize the importance of clear governance mechanisms (such as helmet mandates, speed limits, parking zones, and data privacy standards) in building public trust and ensuring equitable access. The proliferation of policy-focused literature in recent years (Médard de Chardon et al., 2017; Sareen et al., 2021) suggests a need for robust governance frameworks that balance stakeholder interests. These include municipalities, private operators, users, and communities, all of whom play a role in shaping equitable and sustainable shared micromobility systems.
Themes such as “Decision Making,”“Demand Prediction,” and “Trip Choice” are less frequently studied but remain strategically important. Research in these areas supports data-driven planning, capacity forecasting, and service personalization (Castro et al., 2019; Fazio et al., 2021; Folco et al., 2023; Ignaccolo et al., 2022; Reck et al., 2020; Xu et al., 2022; Zhang et al., 2019). These clusters provide methodological foundations for optimizing deployment strategies and understanding modal shift dynamics in urban settings.
Taken together, the 12 thematic clusters identified through LDA reflect a research field that is not only expanding in scope but also diversifying in focus. The field has evolved from a technology- and feasibility-driven phase to one that addresses pressing questions of social acceptance, regulatory adaptation, environmental sustainability, and operational resilience. This multidimensional structure confirms that shared micromobility should not be studied through a single-disciplinary lens but through interdisciplinary frameworks that combine urban planning, behavioral science, public policy, and data analytics.
Emerging Trends in Shared Micromobility Research (RQ3)
The findings reveal a dynamic thematic evolution in shared micromobility research, shaped by shifting societal demands, technological advancements, and growing sustainability concerns. Topics such as “Travel Patterns” and “Technology Acceptance” not only have high volume but also exhibit strong momentum, indicating sustained scholarly attention and increasing practical relevance. This trajectory suggests a convergence between behavioral research and the design of digital micromobility platforms.
In this context, future research could investigate how real-time data, interface usability, and incentive structures influence travel choices, particularly through app-based systems. Such studies can help align user needs with service design, enhancing user retention, and operational efficiency.
The emergence of “Travel Modes and Transportation” as a fast-growing topic highlights increasing interest in multimodal integration. This reflects an evolving research agenda focused on how shared micromobility can complement existing urban transport networks (including public transit, ride-hailing, and active modes). Key research opportunities include interoperability, shared payment systems, and co-located infrastructure designed to support seamless modal transitions.
Conversely, the declining momentum of themes such as “Decision Making” and “System Optimization” may indicate a saturation of early-stage frameworks or methodological redundancy. However, this shift also opens avenues for innovation. Integrating these topics with emerging technologies (such as AI-driven fleet allocation, predictive maintenance, or urban simulation modeling) could revitalize them with methodological and practical relevance.
Although less prominent in volume, the “Demand Prediction” cluster holds strategic significance for improving resource allocation, vehicle distribution, and service responsiveness. Future studies could use real-time usage data and adaptive algorithms to enhance demand forecasting, thereby increasing operational efficiency and minimizing service downtime.
The continued prominence of “Regulations and Safety” and the rising momentum of “Environmental Impact” signal a deepening engagement with normative and policy-oriented concerns. Legal frameworks, equitable access, and climate mitigation have become integral to the shared micromobility discourse. Future research in this area can benefit from interdisciplinary collaborations that incorporate urban law, environmental modeling, public health, and social justice perspectives.
In addressing RQ3 (the progression of trends in shared micromobility research), we present Figure 6, a graphical representation of the intertopic dynamics and thematic relevance within the field from 2010 to 2023. The left side of the figure provides an Intertopic Distance Map (generated via multidimensional scaling) to depict the relative proximity and divergence among identified topics throughout the period. This visual mapping illustrates not only the thematic clustering but also the evolutionary pathways that connect research areas as they expand or contract over time.
Intertopic distance map and keyword relevance in shared micromobility research.
Complementing this spatial analysis, the bar chart specifies the Top 30 most relevant terms for Topic 5 (“Travel Patterns”), revealing the linguistic core of this research topic. The variable shades of blue represent the terms’ frequencies within Topic 5 against their overall corpus distribution, thereby demonstrating the topic-specific lexicon that has become more or less pronounced during the timeframe. The figure thus serves as a tool for understanding the complex and dynamic shifts that have characterized shared micromobility research, providing a clear visualization of the changing focus areas as reflected in RQ3.
The proximity of the topics “Travel Patterns” and “Travel Modes and Transportation” indicates a strong thematic correlation, suggesting that these topics often co-occur in the literature and share a common research focus: how urban mobility patterns influence and are influenced by existing transport modes.
Conversely, topics such as “Innovative Mobility Services” and “Safety” appear more distant on the map, signifying distinct research streams within shared micromobility. Their separation suggests divergent focal points within the field: “Innovative Mobility Services” concentrates on advancements and new models in micromobility offerings, while “Safety” focuses on protocols, regulations, and measures to ensure the safe operation of these services.
According to the adjacent bar graph, in the context of “Travel Patterns,” terms such as “journey,”“metro,” and “built environment” are prominent. This highlights the specificity and high-frequency vocabulary associated with this topic. This visual analysis aids in deconstructing the shared micromobility domain, revealing that topics such as “Regulations” and “Safety” are often discussed in tandem, reflecting the industry’s regulatory emphasis on safe implementation. Meanwhile, the separation between “Environmental Impact” and “Decision Making” may reflect an ongoing dialogue in the literature: specifically, how environmental considerations are integrated into strategic urban planning decisions.
Incorporating these insights into the shared micromobility discourse illuminates the interdependencies among research areas and the distinctive pathways of inquiry prevalent in the field. This structural understanding is invaluable for identifying potential for interdisciplinary research and prioritizing areas for in-depth analysis, thereby facilitating a more cohesive approach to advancing discourse on urban mobility solutions.
Collectively, these data illustrate how shared micromobility research has evolved at the intersection of societal needs and technological innovations, delineating the dynamics of this evolution. The identified trends serve as a critical guide for shaping the direction of future shared micromobility research and understanding its potential impact on urban planning, policymaking, and industrial strategy. Ultimately, this analysis provides a strategic framework that supports more informed decision-making by researchers and policymakers working to advance sustainable and inclusive urban mobility systems.
Conclusion
This study employs Latent Dirichlet Allocation (LDA) to examine the extensive literature on shared micromobility spanning from 2010 to 2023. The findings unveil a multifaceted narrative concerning shared micromobility’s evolving role in urban transport, its intersection with user behavior and policy, and pathways for future optimization. By identifying 12 major thematic clusters (ranging from technology acceptance to fleet management and environmental impact) this research maps the intellectual structure of the field and highlights both dominant and emerging lines of inquiry.
These insights offer a structured foundation for understanding how shared micromobility is positioned within broader urban mobility systems and how its research agenda reflects societal, technological, and environmental shifts. Moreover, this synthesis contributes to the evidence base needed to inform adaptive and responsive policymaking in rapidly evolving urban contexts.
Theoretical Implications
The analysis enriches the theoretical depth of shared micromobility research by demonstrating its convergence with frameworks such as the technology acceptance model (TAM) and Diffusion of Innovation Theory. The prominence of “Technology Acceptance,”“User Behavior,” and “Travel Patterns” as key themes confirms the relevance of these models in understanding the diffusion and societal integration of shared micromobility technologies.
Additionally, this study contributes methodologically by demonstrating how unsupervised machine learning techniques, such as Latent Dirichlet Allocation (LDA), can be effectively used for inductive theory development and comprehensive literature synthesis in interdisciplinary domains. By revealing conceptual relationships and latent structures that traditional systematic reviews might overlook, the analytical rigor of the LDA model not only advances the epistemological foundations of the field but also provides a replicable and transparent means of identifying both dominant and emerging intellectual patterns, thereby advancing methodological standards for future literature synthesis in mobility and urban studies.
Practical Implications
This study provides practical implications for urban planners, policymakers, and shared micromobility service providers seeking to design inclusive, safe, and sustainable transportation ecosystems. The identification of high-momentum topics such as “Technology Acceptance” and “Innovative Mobility Services” highlights the necessity of user-centric design principles and responsive service models. These insights suggest that prioritizing user preferences in the planning process may encourage greater adoption of shared micromobility systems.
In practical terms, this may involve developing multilingual and accessible mobile applications, optimizing pricing structures based on usage data, and designing dynamic interfaces that adapt to user behavior. Additionally, infrastructure investments (such as expanding bike lane networks in high-demand areas or enhancing fleet rebalancing capacity in underserved zones) could support user adoption and satisfaction.
The prominence of “Safety” and “Regulations” underscores the need for robust legal frameworks. These should address helmet mandates, parking enforcement, speed regulations, and user data protection to ensure the responsible integration of shared micromobility into urban mobility systems. Such frameworks are essential not only for public trust and legal compliance but also for greater societal acceptance.
Moreover, findings related to “Environmental Impact” reinforce the potential of shared micromobility systems to contribute to climate action targets by reducing carbon emissions and shifting users away from car dependency. Urban planners can use this potential by aligning shared micromobility initiatives with sustainability and livability goals.
Finally, insights into “Travel Patterns” and “Transportation Modes” provide practical guidance for effective network integration. For example, planners may use these findings to strengthen last-mile connectivity to public transit or inform demand-responsive deployment strategies. This could help reduce congestion, expand multimodal access, and enhance the overall efficiency of urban mobility systems.
From a methodological standpoint, the data-driven insights generated through LDA offer an empirical foundation for these recommendations. Policymakers and shared service providers can therefore base their strategic interventions on objective patterns emerging from the global research landscape, ensuring that shared micromobility policies are both evidence-informed and context-sensitive.
Limitations
The limitations of this study relate to the scope of the dataset and the nature of the LDA method. The size and selection of the sample are significant factors that influence the analysis results. Subjective aspects of LDA modeling (especially the choice of topic number and terms) can significantly impact the outcomes. These factors should be considered in the analysis’s interpretation. Despite these limitations, the probabilistic and unsupervised nature of LDA remains a methodological strength, allowing the extraction of latent semantic structures beyond the reach of traditional content analysis.
Future Research
Building on the findings of this study, future research could pursue several directions. One potential avenue involves applying dynamic topic modeling (DTM) techniques to capture fine-grained temporal shifts in the evolution of shared micromobility research. This approach would enable the identification of transitional themes and emerging discourses influenced by external events, such as the COVID-19 pandemic.
Comparative studies across different geographic regions (particularly between countries with diverse governance structures and infrastructure capacities) could yield insights into localized adaptations and operational challenges of shared micromobility systems. Furthermore, future research could explore how the identified thematic clusters manifest across city types (such as megacities, medium-sized cities, or low-density urban areas) to understand how “Travel Patterns,”“Technology Acceptance,” and “User Demand” interact with the urban context.
In addition, exploring segmented user groups (such as variations in behavior and attitudes by age, gender, or income) could uncover specific adoption barriers and motivating factors, thereby supporting inclusive and equitable service design. Moreover, integrating topic modeling with complementary techniques (such as sentiment analysis or discourse analysis) could enrich understanding of public perception, social resistance, and the cultural narratives shaping shared micromobility adoption.
Collectively, these future research directions hold the potential to deepen theoretical and practical knowledge by promoting context-sensitive, data-driven, and interdisciplinary investigations into the evolving ecosystem of shared micromobility, and advance the methodological integration of machine learning in urban transport studies.
Footnotes
ORCID iDs
