Digital Technology Increases the Sustainability of Cross-Border Agro-Food Supply Chains: A Review

Abstract

The increasing prominence of climate change, geopolitical crises, and global economic slowdown highlights the challenges and structural deficiencies of traditional cross-border agro-food supply chains. As a result, there has been a growing consensus on the need to leverage digital technology to rebuild and innovate a safe, stable, and sustainable global food system. This study assessed the knowledge progress and development trends in the sustainable development of cross-border agro-food supply chains enabled by digital technology. A total of 352 authoritative papers from the core Web of Science database were selected for analysis. The Citespace tool was utilized to visually examine research elements. The findings reveal that research outcomes in this territory experienced a significant period of rapid growth, particularly after 2020. Sustainability and IEEE Access are the journals with the highest and second-highest number of publications. China and the France National Institute are the countries and research institutions with the largest number of publications in this field. The research hotspots are mainly the application of digital technologies, food safety, and supply chain system model innovation. In the past ten years, the research in this territory has gone through three stages: precise timeliness orientation, intelligent strategic decision-making orientation, and model predictability orientation. We further construct the ‘antecedent–practice–performance’ conceptual framework of the sustainability of the digital technology-enabled cross-border agro-food supply chain. Finally, this paper presents the potential research directions in this territory, focusing on four aspects: research method, research mechanism, research topic, and research frontier.

1. Introduction

Given the ongoing progress and evolution of global trade, ensuring the sustainability of cross-border agricultural food supply chains has emerged as crucial issue attracting global attention [1]. As a complex and massive system, a cross-border agricultural food supply chain involves multiple stakeholders [2,3]. Their high-quality and secure operation is crucial to the global food system’s safety and agriculture’s sustainable development [4]. However, faced with the complexity and variability of the global situation, such as the spread of recent outbreaks like the COVID-19 pandemic, highlights the vulnerability of the supply chain and accelerates the demand for digital transformation [5]. Climate change and natural disasters pose a threat to agricultural production and supply chain stability, and digital technologies such as big data are needed to improve crop resilience and optimize resource use [6,7]. Local conflicts and global economic downturns may lead to turbulence in the supply chain and the market. Digital technologies such as the internet of things and big data are needed to monitor risks in real-time and respond quickly to market changes [8], as well as the existing problems of information asymmetry, improving transportation efficiency and risk management in cross-border agricultural food supply chains [9,10]. These challenges have stimulated researchers’ enthusiasm for the application of digital technology in cross-border agricultural supply chains, which in turn has promoted the development of scientific research in this field.

Reviewing the literature to understand research progress in the sustainability of cross-border agricultural food supply chains empowered by digital technologies, the current areas of focus include the following:

(1)

Constructing an intelligent agricultural food supply chain utilizing big data: Cravero et al. (2022) provide insights into the types of data used in agro-food big data, along with their main challenges and trends [11]. Shen et al. (2022) employ big data technology to construct high-performance agro-food recommendation algorithms [12]. Donaghy et al. (2021) discuss the expected applications of big data in dynamic risk management within the agriculture and food supply chains [13].

(2)

Blockchain-enabled traceability in the agricultural food supply chain: Bhat et al. (2022) propose an agro-scm-biot architecture [14]. Srivastava et al. (2023) analyzed the application of blockchain technology in agriculture and food supply chains [15]. Mirabelli and Solina (2020) study the traceability of blockchain in agriculture and food supply chains [16].

(3)

Review studies on digital technology’s empowerment of agricultural food supply chains: Yadav et al. (2022) conduct a systematic review of the application of industry 4.0 technologies in the agriculture and food supply chains [17]. Oliveira and Silva (2023) conducted a review study on the positive impacts, challenges, and future trends of artificial intelligence in the agriculture and food supply chains [18]. Henrichs et al. (2023) present a literature review on the application of digital twins in the food supply chain [19]. Additionally, other scholars have shown that digital technologies such as the internet of things [20,21], machine learning [22,23], and cloud computing [24,25] play important roles in optimizing the traceability, operational efficiency, and risk management of agriculture and food supply chains.

(4)

Conducting sustainability research on cross-border agricultural food supply chains: ① Economies dimension: Barrett, et al. (2022) suggested that the agricultural food value chain stimulates significant direct foreign investment [26]. Gava et al. (2018) discussed using life cycle assessment tools to monitor environmental policy interventions in agricultural food [27]. ② Social dimension: focusing on supporting research on transnational agri-food production networks and proposing corporate social obligation-driven recommendations [28]. ③ Environmental dimension: the emphasis is on evaluating the environmental impact of specific product supply chains, including the apple dessert supply chain [29], milk supply chain [30], integration of environmental policies in the food supply chain [31], and regional agriculture and food supply chains [32]. ④ Resilience and security dimension: the research primarily focuses on the factors influencing security [33,34], studies on security assurance strategies [35], and a resilience assessment of the agricultural food supply chain [36,37].

The existing research significantly contributed to empowering cross-border agricultural food supply chains with digital technologies. However, there are still some limitations and areas that require further investigation. Firstly, the existing review studies mainly rely on a subjective analysis and interpretation of the literature, with a limited number of literature samples and the need for more scientific tools for analysis. Secondly, most of the existing literature review studies in this field emphasize the improvement in production and circulation efficiency and the economic benefits after the application of digital technology and pay less attention to the social and environmental benefits using digital technology to empower the cross-border agricultural supply chain. Thirdly, even if some scientific bibliometric software is used, due to the lack of a theoretical sublimation of the research topic, there is no conceptual framework for related topics. Therefore, the present study selected 352 relevant sample papers from the past decade through a rigorous selection process. By utilizing the scientific literature metrics tool, Citespace, to address the following research questions, visual analysis, and interpretation will be conducted on the sample papers.

Question 1: what is the annual publication volume and trend in this research territory? What are the hotspot journals in terms of publications?

Question 2: what are the hotspot countries, institutions, core authors, and collaborative networks among authors in this research territory?

Question 3: what are the research hotspots and their evolutionary trends in this territory?

Question 4: what is the conceptual framework of digital technology that enables the sustainability of cross-border agricultural product supply chains?

Question 5: what are this territory’s prospective research directions and frontiers?

This paper’s structure and contents are as follows: Section 2 introduces the study approach and data collection process. Section 3 analyzes and discusses the annual publication volume of sample papers and hotspot journals. Section 4 presents the visual results of hotspot countries, institutions, core authors, author collaboration networks, research hotspots, and their evolutionary trends, followed by a discussion and analysis. Section 5 constructs a conceptual framework of a cross-border agricultural product supply chain enabled by digital technology. Section 6 presents the conclusions of this study and highlights its significance and potential research directions.

2. Study Approach and Data Collection Process

2.1. Study Approach

Most of the current literature reviews mainly rely on a subjective interpretation, summarization, and synthesis of the literature in the territory, which is only suitable for literature review studies with many small papers. Citespace 6.2.R3 is a scientific bibliometrics software commonly used for literature reviews, knowledge map construction, and visual analyses of scientific research [38,39]. It aims to help researchers discover the research frontiers, key topics, and academic trends in large-scale literature data [40]. Compared to other literature visualization tools, such as VOSviewer 1.6.20 [41], SciMAT v1.1.04 [42], and Gephi 0.10.0 [43], citespace offers more comprehensive functionality, is more popular, and has higher reliability [44]. These reasons justify our choice of Citespace 6.2.R3 as the analytical tool for this study.

2.2. Data Collection and Processing

This study obtained sample papers from the Web of Science (WoS). WoS provides high-quality, accurate, and reliable academic literature data, including highly influential journals, conference proceedings, books, and other scholarly resources worldwide [45]. It is commonly used for academic evaluations, disciplinary development, and bibliometric analyses [46]. While databases such as SCOPUS, Google Scholar, and Dimensions are also available for selection, this is better-suited to large-scale bibliometric analysis and has higher reliability when combined with the bibliometric software Citespace 6.2.R3 [47]. To ensure the comprehensiveness of the literature sample and its relevance to the research topic, rigorous selection criteria were applied. We formulated the search formula based on the title of this paper. The search formula was divided into four levels. After multiple adjustments and attempts, the search formula for this study was finalized (

Table 1. Search formula statistics table.

Digital technology level	A = (“artificial intelligence*” OR “deep learning” OR “smart system*” OR “neural networks” OR “semantic web” OR “human-computer Interaction*” OR “big data*” OR “ChatGPT” OR “internet of things” OR “5G” OR ”industry 4.0” OR “internet*” OR “digitization*” OR “informatization*” OR “automation*” OR “networking*” OR “blockchain*” OR “machine learning*” OR “data mining*” OR “intelligent*” OR “modernization*” OR “cloud computing*” OR “AR” OR “RFID” OR “AIOT” OR “GPS” OR “GIS”)
Cross-border level	B = (“cross border*” OR “import” OR “global*” OR “world*” OR “international*” OR “multinational*” OR “import and export*” OR “transnational*” OR “foreign trade”)
Agricultural food level	C = (“agriculture*” OR “farm product*” OR “agricultural good*” OR “foodstuff*” OR “agro-food” OR “fresh food” OR “grain*” OR “rice*” OR “wheat*”)
Supply chain level	D = (“logistic*” OR “supply* chain*” OR “transport*” OR “traffic*” OR “supply network*” OR “supplier chain*” OR “circulation*” OR “distribution*” “warehousing*” OR “store*” OR “e-commerce*” OR “e-business*” OR “material flow*”)
Search formula	TI = A and B and C and D
Acronyms note	5G (5th Generation Mobile Communication Technology) ChatGPT (Chat Generative Pre-trained Transformer) AR (Augmented Reality) RFID (Radio Frequency Identification) AIOT (Artificial Intelligence & Internet of Things) GPS (Global Positioning System) GIS (Geographic Information System)

).

In 2014, blockchain technology gained substantial prominence within the cross-border agricultural food supply chain, while digital solutions like cloud computing and big data began to be extensively embraced [48,49]. Therefore, the literature search for this study was conducted from 1 January 2014 to 31 July 2023. The search used the “topic” search rule, resulting in 727 sample papers. After three rounds of screening, a final selection of 352 sample papers was obtained for the review analysis and interpretation of the research topic in this study. The specific process of literature selection is illustrated in Figure 1.

3. Research Output and Key Journals

3.1. Research Output by Year

Analyzing the distribution of literature by year can reflect the research hotspots and attention given to a specific research topic in a particular year, making it a significant research method for analyzing the development trends in a territory [50]. Utilizing the statistical findings, the research output in this territory throughout the last ten years can be categorized into three distinct phases.

The initial phase spanned from 2014 to 2017, characterized by a relatively low level of research attention and output in this territory, with an annual research output of, at most, 15 papers per year. The second stage is from 2018 to 2020, accompanied by the rise of digital technology-enabled convenient modes of cross-border trade in agricultural food and services [51,52]. During this stage, scholars’ enthusiasm for research in this territory rapidly grew, leading to an increased research output. The third stage is from 2021 to 31 July 2023. The research output in this territory reached a stable period characterized by high quantity and quality in the past two years. The research output in 2022 peaked in the last decade, with a significant number of high-quality papers reaching around 100. The research output in 2023 (as of 30 July) reached 41 papers, surpassing the total research output from 2014 to 2017. According to the exponential trend line (dashed line) prediction shown in Figure 2, the projected research output for 2023 is anticipated to range between 80 and 100 papers, indicating a comparatively elevated level of research output over the past decade. The research output regarding the sustainability of the cross-border agricultural food supply chains, empowered by digital technologies exhibited a consistent upward trend, with particularly explosive growth in recent years. It is expected that the number of related papers has increased over time with the advent of new tools.

3.2. Analysis of Research Output in Key Journals

Analyzing the research output in prominent journals is crucial in identifying the focal points within a research territory [53,54]. In this study, the publication frequency of the journals that featured the 352 sample papers was statistically analyzed using the “countif” function in Excel. The findings are presented in

Table 2. High-frequency journal frequency statistics of the sample literature.

Rank	Journal Title	Frequency of Occurrence	Journal Impact Factor/JCR Partition	Subject	Publishing Country
1	Sustainability	24	3.889/Q2	Environmental Science and Ecology	Switzerland
2	IEEE Access	12	3.839/Q2	Engineering Electronics	United States
3	Journal of Cleaner Production	11	11.032/Q1	Environmental Science and Ecology	United States
4	Computers and Electronics in Agriculture	9	8.176/Q1	Agricultural and Forest Science	England
5	Sensors	9	3.874/Q2	Engineering Technology	Switzerland
6	Foods	8	5.088/Q1	Agricultural and Forest Science	Switzerland
7	Agronomy-Basel	6	3.635/Q2	Agricultural and Forest Science	Switzerland
8	Frontiers in Plant Science	6	5.415/Q1	Botanical Sciences	Switzerland
9	Agricultural Systems	5	6.600/Q1	Agricultural Science	Netherlands
10	Environmental Science and Pollution Research	5	5.776/Q1	Environmental Science	Germany

. The subsequent discussion and analysis are presented in the following.

(1) The research output regarding the digital technologies empowering the cross-border agricultural food supply chains exhibits a commendable quality standard. Notably, the top ten pivotal journals account for a substantial portion of the literature, encompassing 95 papers and representing 26.9% of the sample literature. These journals are classified as q1 and q2 in the journal citation reports (JCR). The top two journals regarding research output are Sustainability and IEEE Access.

(2) The countries of publication for the key journals show regional imbalances and differences. The key journals are mainly published in European and American countries such as Switzerland, the United States, the United Kingdom, and Germany. In terms of the number of publications, none of the top 10 journals is distributed in Asian, African, or South American countries. This demonstrates the close connection between the intensity and outcomes of academic research and a country’s economic development. The academic achievements in this area primarily originate from developed countries in Europe and America, indicating a specific regional clustering. Scholars and institutions worldwide should strengthen their international collaboration, and developed countries should provide the necessary technical and methodological guidance and assistance to academic research in developing countries.

(3) The key journals primarily belong to environmental and ecology, agricultural food, and forestry sciences, as well as the areas of engineering and technology. This indicates that the research outcomes in this territory go beyond pursuing advances in digital technologies and improving cross-border supply chain efficiency. These outcomes suggest a desire to protect the environment and ecology at various stages of the supply chain [55].

4. Visual Results and Discussion Analysis

4.1. Hotspot Countries in Terms of Research Output

Examining the hotspot countries in a research territory offers valuable insights into the regional research intensity within that area [56]. In this study, Citespace was employed to analyze the publication countries for the sample literature spanning the last ten years. Additionally, we used the “arcgis 10.8” software to create regional hotspot maps of all publication countries. Figure 3 and Figure 4 show that the first tier of hotspot regions is concentrated in East Asia, South Asia, and North America. Centrality is usually used to measure the importance of a node and the degree of its connection with other nodes. When the node centrality ≥ 0.1, this means that the node has good centrality and is closely related to other nodes [57,58,59]. The country with the highest publication count is China (82 papers), followed by the United States (63 papers) and India (46 papers). The second tier of hotspot regions is mainly concentrated in western European countries, including the United Kingdom (36 papers), Italy (24 papers), Germany (22 papers), and Spain (22 papers). Furthermore, many countries in Africa, South America, and Western Asia still need to publish cutting-edge papers in this territory. The specific discussion and analysis are as follows:

(1) The quantity of research outcomes in this area closely relates to a country’s economic development and arable land area. The United States ranks first regarding GDP and arable land area globally, while China ranks second and third [60]. Arable land provides the foundation for developing a country’s agricultural supply chain. Economic development provides a broad consumer market and sufficient financial support to empower agricultural food and services with digital technologies. This has motivated countries like the United States, China, and India to actively research this territory, driving innovation and the development of digital agriculture and food supply chains both domestically and globally.

(2) Although China has the highest publication count, its centrality is only 0.09, indicating weaker research collaborations with other countries. Regarding the limited research outcomes from most African countries, countries with advantages should strengthen their academic exchange platforms and provide relevant support regarding digital technologies.

4.2. Hotspot Institutions in Terms of Research Output

The quantity of publications reflects an institution’s strength and influence in terms of their research [61]. After analysis, the 352 sample papers in this study were distributed among 227 research institutions, totaling 331 collaboration links. Figure 5 shows the collaboration network of research institutions with three or more publications. Additionally, Figure 6 presents the top 10 research institutions based on publication counts. The discussion and analysis of hotspot institutions are as follows:

(1) Hotspot institutions are closely related to hotspot countries. Generally, hotspot institutions are concentrated in hotspot countries. In this research territory, the top 10 institutions in terms of distribution counts are mainly located in the hotspot countries of North America, East Asia, and Western Europe. Three research collaboration groups emerged in this territory. In terms of scale, the leading group is centered around INRAE, followed by groups centered around the United States department of agriculture (USDA), and the Chinese academy of sciences. These core institutional collaboration groups are also distributed among hotspot institutions and countries.

(2) It is worth noting that the collaboration relationships among hotspot institutions in this territory are relatively loose. Institutions such as the university of California System, Beijing Technology & Business University, and N8 research partnership rank among the top 10 in terms of publication count but have centrality values of 0, indicating weak collaborations with other institutions. Additionally, the collaborative efforts among institutions such as the Chinese Academy of Sciences, Beijing Technology & Business University, and the Beijing Institute of Fashion Technology primarily focus on activities within China, lacking exchanges and collaborations with research institutions in other countries.

4.3. Core Authors and Author Collaboration Network

The author collaboration network reflects writers’ influence in the territory and their collaborative relationships, providing valuable insights into the connections and impact within the research community [62]. In this study, the author collaboration network consists of N = 266 nodes and E = 317 links, with a network density of D = 0.009. The authors with the most publications are Zhang, Lin, and Xu, Jiping, with six papers. Figure 7 shows the author collaboration network, where certain author collaboration groups have emerged. For example, there is an academic group represented by Zhang, Lin (six papers), Xu, Jiping (six papers), and Zhao, Zhiyao (five papers), with publications concentrated in 2022–2023. Their research focuses on the role of blockchain technology in improving the efficiency and sustainability of the agricultural food chain [63,64]. Another academic group represented by Hu, Xiangpei (two papers) and Ruan, Junhu (two papers) has publications concentrated in 2019; their research primarily concentrates on the financial applications of the internet of things (IoT) and the involvement of the IoT at the beginning of the agriculture and food supply chains [65,66]. Additionally, Cinar and Ilkay (two papers) and Koklu and Murat (two papers) cooperate to study how machine learning methods can be useful in different parts of the agricultural industry [67,68].

The central creators in this area have made substantial contributions and are the main thrust behind the research and its pragmatic application [69]. Price’s Law is an important method to identify the core authors in a specific area. It states that, within a topic, papers are established by a group of beneficial creators; furthermore, the size of this creator bundle is generally identical to the square groundwork of the total number of creators [70]. The minimum number of papers that the core authors should have published, as shown in Equation (1), is M. Nmax. This equation addresses the number of papers by the creator with the most noteworthy output included in the area.

$$\mathrm{M}\approx 0.749\sqrt{\mathrm{Nma}}\mathrm{x}$$

(1)

According to

Table 3. Statistics of core authors and published papers in the territory.

Rank	Author	Number of Publications	Rank	Author	Number of Publications
1	Zhang, Xin	6	13	Peng, Xiangzhen	2
2	Xu, Jiping	6	14	Cropotova, Janna	2
3	Zhao, Zhiyao	5	15	Hu, Xiangpei	2
4	Wang, Xiaoyi	5	16	Chan, Felix Tung Sun	2
5	Li, Haisheng	3	17	Baeumner, Antje J	2
6	Shi, Yan	2	18	Figorilli, Simone	2
7	Hassoun, Abdo	2	19	Antonucci, Francesca	2
8	Qi, Zhibo	2	20	Pallottino, Federico	2
9	Cinar, Ilkay	2	21	Koklu, Murat	2
10	Li, Qingquan	2	22	Barba, Francisco J	2
11	Jagtap, Sandeep	2	23	Ruan, Junhu	2
12	Costa, Corrado	2

, the creator with the most noteworthy output in the area has contributed to six papers. Therefore, Nmax = 6. Calculating Equation (1), we find that “M” = 1.835, demonstrating that creators who have contributed to at least two papers are viewed as the central creators in this area. The core authors in this area are listed in Table 2. Price’s Law suggests that when the cumulative publication count of the core authors reaches or exceeds 50% of the total publication count, this indicates the formation of the core author group in that area [71]. In this case, the cumulative publication count of the core authors in this area is 57 papers, accounting for 16.2% of the total publication count. This rate is far below half of the total, suggesting that the region has not yet developed a group of central creators.

4.4. High-Frequency Keywords and Research Hotspots

A research hotspot refers to an interconnected group of scientific topics explored in a relatively large number of papers during a specific period [72]. High-recurrence catchphrases play a fundamental role in summing up the principal subjects of the articles and are a pivotal device in breaking down problem areas of exploration [73,74]. Therefore, in this study, an analysis was conducted of 352 sample articles to identify high-frequency keyword co-occurrence patterns, as shown in Figure 8.

Equation (2), proposed by Donohue in 1973 and obtained via ZIPF’s Subsequent Regulation, characterizes high- and low-recurrence words [75]. In Equation (2), T refers to the minimum requirements for high-recurrence catchphrases, and I₁ alludes to the number of watchwords that showed up only once in the information. In this study, I₁ = 122; thus, T ≈ 15.13. Hence, words that appear 16 times or more are defined as high-frequency keywords in this field, as displayed in

Table 4. High-frequency keywords and centrality statistics in the territory.

Rank	Keyword	Frequency of Occurrence	Centrality	Rank	Keyword	Frequency of Occurrence	Centrality
1	Big data	51	0.19	10	system	24	0.04
2	Supply chain	40	0.12	11	systems	20	0.02
3	Technology	39	0.05	12	food security	19	0.05
4	Management	39	0.08	13	artificial intelligence	19	0.06
5	Internet	31	0.04	14	challenges	18	0.05
6	Machine learning	29	0.05	15	Internet of things	18	0.09
7	Agriculture	28	0.14	16	quality	16	0.02
8	Precision agriculture	25	0.14	17	traceability	16	0.03
9	Model	25	0.11	18	impact	16	0.07

.

$$\mathrm{T}= \frac{1}{2}(-1+\sqrt{\left(1+8{\mathrm{I}}_1\right)})$$

(2)

Keyword clustering is very important for revealing hot topics in the research field [76]. This study used the log-likelihood ratio (LLR) algorithm for keyword clustering. As shown in Figure 9, the analysis resulted in seven significant clusters of keywords, where the size values represent the number of keywords within each cluster [77]. The clustering modularity value “Q” was calculated to be 0.4756, more significant than 0.3, representing a vital clustering structure. The typical outline value “S” is 0.7786, surpassing 0.7, exhibiting a convincing clustering result and strong data for the following discussion and focusing on important areas of investigation [78]. By summarizing and categorizing the high-frequency keywords and clusters, the following three major research hotspot themes in this territory can be identified, as shown in

Table 5. Domain clustering summary classification table.

Research Theme	Cluster Label	Cluster Size	High-Frequency Keywords
The application of main digital technology categories in the territory	#0 deep learning #1 internet of things #4 data mining #6 Blockchain	41 34 30 22	big data; machine learning; artificial intelligence; technology; smart farming; precision agriculture; convolutional neural network; digital transformation; network formation; autonomous robots;
Digital technology enables sustainability and food safety in cross-border agricultural food supply chains.	#2 food security #3 climate change	34 32	quality; traceability; challenges; sustainable development goals; decision support; digital technologies; technological innovation; rainfall prediction; cold chain management; postharvest quality; crop modelling;
System model innovation of cross-border agricultural foods supply chain enabled by digital technology	#5 supply chain management	24	management innovation; model; system; impact; land use modeling; efficiency improvement; simulation games; agricultural foodstuff system;

. The results and discussions are presented as follows.

The first research theme focuses on applying digital technologies to the sustainability of cross-border agricultural food supply chains. This theme includes clusters such as #0 deep learning (41), #1 internet of things (34), #4 data mining (30), and #6 blockchain (32). Critical keywords within this theme include big data, machine learning, artificial intelligence, technology, and smart farming. These studies highlight the mainstream trends in digital technologies for cross-border agricultural food supply chain applications. Mainstream technologies like deep learning, blockchain, data mining, and big data continuously optimize the cross-border agricultural food supply chain structure, enhancing operational efficiency at various nodes. Ancin et al. (2022) explore the positive impact of artificial intelligence, big data, and cloud computing on production efficiency, crop yields, and cost reductions in the supply chain [79]. Klerkx et al. (2019) provide new insights into the connection between digital agricultural food technologies [80]. Jararweh et al. (2023) propose that integrating intelligent technologies can overcome bottlenecks in the agriculture and food supply chains by reducing the need for manual interaction and enabling proactive, intelligent decision-making [81].

The second research theme focuses on the sustainability and food security aspects of digitizing technology-enabled cross-border agricultural food supply chains. This theme includes clusters such as #2 food security (34) and #3 climate change (32). Critical keywords within this theme include quality, traceability, challenges, sustainable development goals, and decision support. The global cross-border trade of agricultural foods is facing difficulties and challenges due to natural disasters, climate change, and escalating conflicts, leading to heightened concerns about food safety among people worldwide [82]. Chen et al. (2021) proposed a secure, reliable, and convenient transportable and traceable agricultural food mechanism model by incorporating IoT technology [83]. Ping et al. (2018) studied the broad utilization of advancements like radio recurrence recognizable proof (RFID) and remote sensor organizations (WSN) to guarantee the security of the agricultural food store network [84]. Guruswamy et al. (2022) discuss the role of digital technologies like the internet of things, 6G, and artificial intelligence in enhancing the security of global foodstuff systems and supply chains [85].

The third research theme focuses on innovating system and model of the cross-border agricultural food supply chain in the context of digital technologies. This theme includes cluster #5, supply chain management (24). Critical keywords within this theme include management innovation, model, system, impact, land use modeling, and efficiency improvements. The applicability of traditional supply chain management strategies to digital supply chain management has remained strong via the incorporation of brand-new digital technologies into the long-term viability of the cross-border agricultural food supply chain. Therefore, scholars have focused on innovating management models, methods, and system efficiency to achieve digital technology improvements. Integrating digital technologies into supply chain management reduces labor costs and improves supply chain efficiency. Liu et al. (2012) found that, via the utilization of 5G IoT technology, the supply chain can provide more information about the development of an intelligent supply chain [86]. Peng et al. (2022) propose a multi-blockchain-based fine-grained supervision model for the rice supply chain, thus improving management efficiency [87].

4.5. Burst Keywords and Evolutionary Trends of Research Hotspots

The keyword time zone map represents the transient advancement of exploration areas of interest and uncovers the transformative patterns in various improvement stages [88]. As shown in Figure 10, the time zone map displays the top three keywords in terms of frequency for each year during the research period. “Bursting keywords” allude to watchwords that experience a huge increase in recurrence within a specific period, addressing the unexpected increase in research interest within that period [89]. Figure 11 presents 11 bursting keywords in this research territory, along with the intensity and duration of their outbreak. By combining the timezone maps of research topics at different development stages with bursting keywords, we can analyze the evolutionary trajectory of the research hotspots. This approach allows for the mutual corroboration and supplementation of research findings, ensuring the scientific rigor and effectiveness of this studies [90]. The authors grouped the areas of interest in this study area during the review timeframe into three phases based on the organization, induction, differentiation, understanding, and classification of the keyword timezone map and bursting keywords.

Phase 1 (2014–2017) was the stage of providing the cross-border agricultural food supply chain with timely and precise guidance through digital technology. Keywords and bursting keywords included “agri-foodivity”, “precision agriculture”, “artificial neural network”, etc. From 2014 to 2017, digital technology was crucial in the development of a cross-border agricultural food supply chain. Numerous agricultural food organizations and partners in the supply chain enterprise network started using computerized advances like the internet of things to track the continuous status of agricultural foods throughout the creation, processing, transportation, and sales processes. This provided real-time data and information to the participants, thereby improving the timeliness and precision of the agriculture and food supply chains. Aung et al. (2014) proposed that a traceability system for agriculture and food supply chains, shaped by digital technologies such as blockchain, could ensure supply chain information transfer timeliness [91]. Deichmann et al. (2016) introduced the role of digital technology in promoting the sustainability of cross-border agricultural food supply chains by supplementing other production factors to improve the efficiency [92]. Jayaraman et al. (2016) discussed how internet of things technology could enhance the understanding and prediction of crop performance under various environmental conditions [93].

Phase 2 (2018–2020) was the stage of empowering cross-border agricultural food supply chain sustainability with intelligence and strategic decision-making guidance through digital technology. The keywords and bursting keywords include “big data analytics”, “smart farming”, “decision support system”, etc. From 2018 to 2020, countries worldwide began elevating the application of digital technology in cross-border agricultural food supply chains to the strategy level. The European Union launched a digital agriculture strategy to improve agricultural food efficiency, intelligence, sustainability, and other related processes. In the same year, China introduced the “internet plus agriculture” strategy to promote the digital technology empowerment of cross-border agricultural food supply chains. Chai et al. (2019) proposed that “internet plus” technology could become an essential means of constructing cross-border agricultural food supply chain systems for fresh products, extensively promoting their development [94]. Zhai et al. (2020) discussed how information technology and internet-supported agricultural food strategic decision-support systems could be collected and analyzed [95]. Sinha et al. (2019) argued that the internet of things (IoT) is a crucial driving force for realizing the vision of intelligent agriculture, proposing a user-centric IoT architecture that optimizes the agriculture and food supply chains [96].

Phase 3 (2021–2023) is characterized by modeling and predictive guidance provided by digital technology. The keywords and bursting keywords include “predictive models”, “simulation”, “sustainable”, etc. From 2021 to 2023, research on the use of digital technology to improve cross-border agricultural food supply chain sustainability has increasingly focused on utilizing cutting-edge digital technologies such as AI and 6G networks to provide more accurate and real-time data analysis and predictive capabilities to the participants in various supply chain stages. During this phase, the emphasis shifted towards utilizing advanced digital technologies to enhance decision-making, anticipate future trends, and optimize the cross-border agricultural food supply chains. For instance, Mahroof et al. (2021) employed the ISM method and combined it with the digital technologies of Agriculture 4.0 to model and analyze challenges in the agriculture and food supply chains [97]. Ryan et al. (2023) highlighted the potential of digital technology, particularly AI, in the global digital transformation of cross-border agricultural food supply chains [98]. Hu (2023) utilized the gray prediction model to forecast the increasing demand trends for fresh agricultural foods [99].

5. The Conceptual Framework of Digital Technology Enabling the Creation of a Cross-Border Agricultural Product Supply Chain

5.1. Interpretation of the Connotations of Digital Technology Enabling the Creation of a Cross-Border Agricultural Product Supply Chain

Digital technology can empower the cross-border agricultural product supply chain. This refers to the deep integration of digital technologies, such as big data, the internet of things, blockchain, and artificial intelligence, with all aspects of the traditional cross-border agricultural product supply chain to promote transparent management, rapid response, seamless connection, efficient information sharing, and the collaboration of the cross-border agricultural product supply chain throughout the whole process, to realize the intelligent transformation of the global food system from agricultural products to the food dining tables, thus promoting the construction of a cross-border agricultural product supply chain ecosystem characterized by digitization and networking [100]. The essence using of digital technology to establish this cross-border agricultural product supply chain is to integrate the logistics, capital flow, business flow, information flow, and business processes of a global supply chain system to build a complete intelligent ecosystem. It aims to promote the integration and efficient use of resources and to introduce intelligent decision-making to fundamentally reshape the traditional cross-border agricultural supply chain system’s circulation, storage, sales, and other links.

5.2. The Path and Effect Analysis of Digital Technology Empowering the Cross-Border Agricultural Product Supply Chain

High-quality literature review research should not only use scientific and objective methods to conduct a bibliometric analysis of the sample literature, but also to study the content of the sample literature in depth and summarize the content, as well as when using digital technology. The research should determine which links of the cross-border agricultural product supply chain are empowered and what effect has been achieved. Digital technology was used in the interpretation of the cross-border agricultural product supply chain, in the above literature measurement analysis, and to determine the discussion content. As shown in

Table 6. Effects’ analysis table of the use of digital technology to empower cross-border agricultural product supply chain modes.

Rank	Title and Author	Digital Technology	Empowerment Methods and Effects
1	Data Type and Data Sources for Agricultural Big Data and Machine Learning [11].	big data; machine learning	Uses machine learning and big data to judge climate trends and determine the most suitable time for agricultural planting in the production process.
2	Construction of intelligent supply chain system of agri-foods based on big data [12].	big data; internet of things	Uses big data technology to build a model to solve the problems of production and sales links in the supply chain of agricultural products.
3	Big data impacting dynamic food safety risk management in the food chain [13].	big data	From the perspective of microbial safety, the expected application of big data technology in the dynamic risk management of agricultural product supply chains was discussed, and the risk of microbial infection in agricultural products was identified in real-time during the production process.
4	The Application of Artificial Intelligence and Big Data in the Food Industry [101].	big data; artificial intelligence	Artificial intelligence (ai) and big data technology a become the key to strengthening the safety of the agricultural product supply chain and the efficient promotion of production and marketing links.
5	Spatially and temporally disparate data in systems agriculture: Issues and prospective solutions [102].	big data; cloud computing	A cross-space and time resolution data management system for cross-border agricultural product supply chains was developed using big data technology and cloud computing to ensure the intensive production and diversification of circulation links in the agricultural product supply chain and to enhance the flexibility of the supply chain.
6	Big data cloud computing framework for low carbon supplier selection in the beef supply chain [103].	big data; cloud computing	A new framework based on big data cloud computing technology was developed to achieve the best balance between supply chain operation factors and carbon footprint, to select the most suitable suppliers, and to reduce the carbon footprint of multiple links in the supply chain.
7	The fourth industrial revolution in the food industry-Part I: Industry 4.0 technologies [104].	artificial intelligence; big data; blockchain; internet of things	It is proposed that although every technology, such as big data, internet of things, blockchain, and artificial intelligence, is important, only by combining multiple technologies at the same time can a breakthrough be achieved in the sustainability of cross-border agricultural product supply chains.
8	Big GIS analytics framework for agriculture supply chains: A literature review identifying the current trends and future perspectives [105].	big data; gis	This study proposes a big gis analytics (BGA) framework for cross-border agricultural supply chains. Big data technology is used to improve the quality of agricultural GIS applications to improve agricultural productivity.
9	Agriculture-food supply chain management based on blockchain and IoT: a narrative on enterprise blockchain interoperability [14].	blockchain; internet of things	Agro-scm-biot (agricultural supply chain management using blockchain and internet of things) architecture is proposed to solve the problems of scalability optimization, interoperability, security, and privacy in cross-border agricultural supply chain information systems.
10	Blockchain technology and its applications in agriculture and supply chain management: a retrospective overview and analysis [15].	blockchain	It is proved that the electricity used in the agricultural product supply chain process requires blockchain technology. The sustainable e-agriculture of blockchain technology has brought great convenience to farmers’ sales links and increased the sales of cross-border agricultural products by 20%.
11	Blockchain and agricultural supply chains traceability: Research trends and future challenges [16].	blockchain	The research discusses the traceability of blockchain in the cross-border agricultural product supply chain. The information of each link is fed back to consumers in real-time to promote the construction of a nutritious, safe, and healthy global food system.
12	Blockchain Technology for Transparency in Agri-Food Supply Chain: Use Cases, Limitations, and Future Directions [106].	blockchain	Three areas in which blockchain can be implemented in the cross-border agricultural product supply chain are determined: the distribution of cross-border agricultural products, the sourcing and procurement of agricultural products, and the safety and quality of agricultural products.
13	Research on the Construction of Grain Food Multi-Chain Blockchain Based on Zero-Knowledge Proof [107].	blockchain	Blockchain technology is applied to the cross-border grain supply chain to improve the data trust protection and information coordination consensus of the supply chain.
14	Blockchain-Based Agri-Food Supply Chain: A Complete Solution [108].	blockchain	Based on the weak links in traditional cross-border agricultural product supply chains, such as a lack of transparency, lack of a related responsibility system, and lack of statistical convenience, the functions of blockchain and smart contract are used to enhance the traceability, trust, and delivery mechanism of the whole cross-border agricultural product supply chain.
15	A life cycle framework of green IoT-based agriculture and its finance, operation, and management issues [67].	internet of things	A green internet of things system is built for the whole cross-border agricultural product supply chain, and the problems of internet of things financing and internet of things data management are solved according to the actual situation. The internet of things technology is used to create agricultural production modes and new agricultural comprehensive enterprises.
16	IoT-Based Real-Time Crop Drying and Storage Monitoring System [109].	internet of things	The internet of things technology is used in the production of the agricultural product supply chain; the intelligent system monitors the temperature, humidity, and light during the real-time drying process of corn to ensure the quality and safety of agricultural products.
17	Internet of Things Platform for Smart Farming: Experiences and Lessons Learnt [94].	internet of things	Based on the internet of things, the smart farm net platform is designed to automatically collect and analyze environmental data such as soil, humidity, and temperature, and provide reasonable suggestions. Crop performance is enhanced in the production of cross-border agricultural product supply chains and farm productivity is increased.
18	Systematic review of Internet of Things in smart farming [110].	internet of things	The internet of things (IoT) technology is applied to agriculture, often referred to as intelligent agriculture. Through the agricultural monitoring and control system, automatic irrigation system, and plant disease monitoring system of the internet of things, agricultural production is controlled and managed, and the resource utilization rate of each link in the agricultural product supply chain is improved.
19	Smart Agriculture Cloud Using AI Based Techniques [111].	internet of things; artificial intelligence; machine learning	A cloud-based intelligent system is proposed to adapt to the scenario of remote monitoring and the internet of things (IoT) is used in the cross-border agricultural product supply chain.
20	Applications of internet of things (IoT) and sensors technology to increase food security and agricultural Sustainability: Benefits and challenges [112].	internet of things	The potential applications of IoT and sensor technologies in cross-border agricultural supply chains are studied, including in irrigation monitoring systems, crop disease detection, monitoring, forecasting, and harvesting, and climate condition monitoring.
21	A systematic literature review on machine learning applications for sustainable agriculture supply chain performance [22].	machine learning	According to the application of the machine learning algorithm in different links in the cross-border agricultural product supply chain, an application framework based on machine learning is proposed to improve the productivity of the production links in the cross-border agricultural product supply chain and the sustainability of each link in the whole process.
22	A novel artificial multiple intelligence system (AMIS) for agricultural product Transborder logistics network Design in the Greater Mekong Subregion (GMS) [113].	artificial intelligence	The research uses mixed-integer programming and heuristic algorithms to design a supply chain logistics network for the cross-border export of agricultural products in major countries in Southeast Asia and to maximize the profit of the total agricultural product export supply chain.
23	Exploration of cross-border e-commerce and its logistics supply chain innovation and development path for agricultural exports based on deep learning [114].	artificial intelligence; deep learning	The research uses the ‘prophet’ algorithm under deep learning to construct a price prediction model for the cross-border export of agricultural products and analyzes the logistical risks of the cross-border agricultural product supply chain.
24	Hopfield artificial neural network-based optimization method for selecting nodes of fresh agricultural products international logistics network [100].	artificial intelligence	This paper uses the Hopfield artificial neural network to study the model of node location of the international coordination network of fresh agricultural products, to shorten the circulation distance of each node of cross-border agricultural products, improve the efficiency of cross-border agricultural products supply chain circulation, and to save logistics resources.
25	Application of Deep Learning to Production Forecasting in Intelligent Agricultural Product Supply Chain [115].	artificial intelligence; deep learning; internet of things	The prediction of agricultural product output is an important factor in the realization of the sustainability of the cross-border agricultural product supply chain, and the internet of things technology is a prerequisite for accurate prediction of agricultural product output. Based on big data and deep learning technology, the prediction model of agricultural product yield is constructed, and the prediction results are more accurate, which is conducive to the realization of the sustainability of the cross-border agricultural product supply chain.

, this study fully studied the representative sample literature on the efficient empowerment of cross-border agricultural product supply chains with big data, internet of things, blockchain, and artificial intelligence, and created statistics and analyses of the title author, the digital technology that was used, and the means and the effects of this empowerment.

5.3. A Conceptual Framework of Digital Technology Enabling the Creation of a Cross-Border Agricultural Product Supply Chain

In the previous section, the authors, countries, research hotspots, and evolution trends of the literature in this field were analyzed and discussed using the research method of combining bibliometrics with a visual knowledge map. An analysis of the path and effects of the use of digital technology to empower cross-border agricultural product supply chains shows the specific path by which digital technology can empower cross-border agricultural product supply chains and what effects this has achieved. Based on the research results and these two aspects, as shown in Figure 12, we further construct the ‘antecedent–practice–performance’ conceptual framework of digital technology-enabled cross-border agricultural supply chain sustainability. The conceptual framework consists of three core parts.

The first part is the digital management platform of the cross-border agricultural product supply chain at the ‘antecedent’ level. At this level, digital technologies, such as big data, blockchain, IoT, cloud computing, and artificial intelligence, together constitute the core driving force of the digital management platform of a cross-border agricultural product supply chain. The platform not only provides technical support for the sustainability of the supply chain but also provides a scientific basis for the optimization and decision-making of the supply chain through data collection, processing, and analysis.

The second part is the ‘practice’ level of the conceptual framework, which shows the process by which digital technology can empower all aspects of the cross-border agricultural product supply chain. The empowerment of big data technology in the cross-border agricultural product supply chain is mainly reflected in the production, picking, processing, sales, and traceability. In the production process, big data technology provides farmers with scientific production plan recommendations based on meteorological data, soil conditions, and crop growth. The empowerment of IoT in the cross-border agricultural product supply chain is mainly reflected in the equipment management and production process monitoring. The empowerment of blockchain technology in the cross-border agricultural product supply chain is mainly reflected in its traceability and quality assurance. By collecting information on the production, processing, transportation, and sales of agricultural products in the system in real-time, the transparency of the information can be ensured to improve consumers’ trust in the products. The use of artificial intelligence technology to empower cross-border agricultural product supply chains is mainly reflected in the processing and circulation in the chain. In the processing link, automation equipment is used to improve the efficiency and quality of agricultural product processing. In the circulation link, through machine learning, artificial neural networks, and other algorithms, according to the international situation, the cross-border logistics policy is established and the real-time optimization of the import and export of agricultural products along the logistics path is ensured.

The third part is the ‘performance’ level of the conceptual framework, which shows the effects of a cross-border agricultural product supply chain empowered by digital technology. After the traditional cross-border agricultural product supply chain was empowered by digital technology, the efficient interaction, accurate matching, multi-end integration, and intelligent processing of each link were ensured, the whole process of the supply chain was improved, five-in-one big data analysis was established, and a management platform was created for the logistics, business flow, capital flow, information flow, and sales channels of cross-border agricultural product supply chains. The blocking points of the traditional cross-border agricultural product supply chain were used to form a supply chain ecosystem that uses the whole process and all factors.

6. Conclusions and Outlook

6.1. Research Conclusions

This study analyzed and discussed the relevant literature on digital technology-enabled cross-border agricultural food supply chain sustainability from 2014 to 30 July 2023. The research conclusions are as follows:

(1)

From 2014 to 2018, the annual volume of publications related to the digital technology-enabled cross-border agricultural food supply chain sustainability could have been higher, showing a slow growth trend but never exceeding 30 articles per year. After 2019, the research output experienced exponential growth, reaching 100 articles per year in 2022. This indicates that this territory has attracted increasing attention and research from scholars.

(2)

The 352 sample articles in this study were distributed among 77 countries and regions. The first tier of publishing regions is concentrated in East Asia, South Asia, and North America. The countries with the highest number of publications are China (82 articles), followed by the United States (63 articles) and India (46 articles). The second tier of publishing regions is mainly concentrated in Western European countries, represented by the United Kingdom (36 articles), Italy (24 articles), and Germany (22 articles). It is worth noting that many countries in Africa, South America, and Western Asia have yet to publish cutting-edge papers in this area.

(3)

The 352 sample articles in this study were distributed among 227 research institutions, totaling 331 cooperation relationships. INRAE (the National Research Institute for Agriculture, Food and Environment) ranks first, with 10 articles, followed by the University of California System (8 articles) and the Chinese Academy of Sciences (6 articles).

(4)

Zhang, Lin, and Xu, Jiping, with six articles, have the highest number of publications in this territory. Collaborative research groups of authors have emerged in this territory, such as academic groups represented by Zhang, Lin (six articles), Xu, Jiping, Zhao (six articles), Zhiyao (five articles), etc. However, this research area has yet to form a clear core group of authors.

(5)

The ten journals with the highest increase in publications in this research area are all in the q1 and q2 groupings of JCR, meaning that the quality and level of investigation are high. Sustainability, food, agriculture, and computer science are the primary focus areas of the journals with the most publications. This, in a roundabout way, suggests that the direction of exploration in this area is primarily centered around the manageable coordination of cross-border, agricultural food storage networks with the assistance of PC-based computerized advances concerning the climate, society, and economy.

(6)

The areas of interest in this research area predominantly relate to the utilization of improvements in the manageability of cross-border agricultural and food supply chains, the use of computerized innovation-empowered cross-border agricultural food production networks, and the fundamental development of cross-border agricultural and food supply chains using computerized innovations. Over the last decade, the research focus in this area has shown three phases: the exact idealness direction, keen vital dynamic direction, and model consistency direction.

(7)

The digital technology-enabled cross-border agricultural product supply chain sustainability conceptual framework consists of three core parts. The first part is a digital management platform of the cross-border agricultural product supply chain at the ‘antecedents’ level. The second part is the ‘practice’ level of the conceptual framework, which shows the process by which digital technology empowers all aspects of the cross-border agricultural product supply chain. The third part is the ‘performance’ level of the conceptual framework, which shows the effects of a cross-border agricultural product supply chain empowered by digital technology.

6.2. Research Significance

6.2.1. Theoretical Significance

Firstly, a bibliometric analysis and sustainability analysis were conducted: the sample literature included in this study contained 727 sample documents at four levels, and through a layer-by-layer screening, only 327 sample documents related to the research topic ‘digital technology empowers cross-border agricultural supply chain sustainability’ were retained. This reveals the number of publications, countries, institutions, research hotspots, and evolution in this field, and can be used to analyze the development of academic research results in this field, which is conducive to the cooperation of countries and institutions working in this field and promotes the sustainable development of and innovations in this field.

Second, in terms of the sustainable application of digital technology in cross-border agricultural product supply chains, this paper discusses the application of digital technology, such as big data, blockchain, and the internet of things, to improve the transparency, efficiency, risk management, and responses of the supply chain, which are key factors in realizing the sustainable development of the economy, society, and environment of the cross-border agricultural product supply chain.

Thirdly, regarding the conceptual framework and sustainability aspects proposed in the study, the ‘practice’ level shows the application of digital technology in all aspects of the supply chain, and the ‘impact’ level reflects the specific contribution of these technology applications to supply chain sustainability.

Fourth, regarding future research directions and future challenges, the research finally puts forward future research directions in this field, focusing on three aspects—the research approach, research topics, and research limits—and expounds on the important future directions in the sustainable development of cross-border agricultural product supply chains. Five challenges in future research are proposed, such as digital infrastructure construction and narrowing the digital divide, which are prerequisites for ensuring the sustainability of the supply chain.

6.2.2. Practical Significance

First, for scholars in this field, this study can provide them with basic information such as the authors, countries, and institutions focusing on hot topics in this research field, and provide a reference for scholars seeking cooperative innovation. Additionally, this study shows the research hotspots and hotspot evolution in this field and proposes future research directions and challenges, which can help scholars in this research field gain insight into the hot research topics in this field. They can use the future research directions provided by this study to pursue frontier hotspots and promote sustainable innovation in this field.

Secondly, for the enterprises and practitioners interested in cross-border agricultural product supply chains and the research, development, and application fields of digital technology in this area: the hot publishing institutions and authors, countries, and research hotspots shown in this study can be used to ensure the quality and efficiency of each link in the enterprise supply chain. Those working on the research and development of digital technology can refer to the literature referenced in this study focusing on the use of digital technology in cross-border agricultural product supply chains. This study can provide a certain degree of guidance in their research and development of new digital technology that can be applied to cross-border agricultural product supply chains. In addition, the study also talked about sustainable progress, reminding businesses and practitioners to pay more attention to the coordinated sustainability of the environment, society, and economy.

6.3. Research Prospect Directions

Based on the discussion and analysis of the research process and research conclusions, this study proposes future study directions, focusing on four aspects: the study approach, study regime, study topics, and study edge. It provides a broad vision and a variety of possible future research directions.

(1)

Study approach: moving from a solitary examination technique to a mix of various exploration strategies.

While a single research method can provide a concise analysis of research topics, it may need to be revised to ensure comprehensiveness and reduce the inherent flaws of a single method. Currently, research in the area of digitally enabled sustainable cross-border agricultural food supply chains often relies on single methods or subjective analyses, such as using a single method such as a life-cycle assessment (LCA) [116,117] or multi-regional input–output (MRIO) model [118,119]. In future research, it will be essential to adopt multiple research methods to leverage the advantages of different approaches, reduce their limitations, and enhance the scientific rigor and richness of the research. Markov chains, Bayesian networks, and optimization models can all be used in research, particularly when combined with case study methods and multi-objective and two-stage stochastic programming models.

(2)

Study Regime: leveraging the cluster effect of national, institutional, and author collaborations and strengthening innovative collaborative mechanisms.

In light of the statistical results regarding the joint efforts of different countries, institutions, and creators, it is evident that some cooperation has been established in this area. However, a stable cooperative group of countries, institutions, and authors has not yet been formed. Hence, future research in this area should improve the creative, cooperative components and investigate cooperative exploration approaches involving different businesses and locales. Universities, governments, and relevant social organizations should be encouraged to carry out cross-regional and cross-country cooperation and exchanges.

(3)

Study topics: deepening technological innovation, model-driven approaches, and food safety in cross-border agricultural food supply chains.

The global food security summit was held in the UK on 20 November 2023. The summit aimed to explore cutting-edge scientific and innovation achievements and identified technological innovation as a critical direction in the strategic framework of food and agriculture for the next decade. Combining the research hotspots and trends identified in this study, “technological innovation” “artificial intelligence” and “model-driven approaches” emerge as bursting research topics in the territory. Therefore, the research prospectus should focus on the development of suitable digital technologies and application models, including the application of artificial intelligence, to improve the efficiency of supply chain systems, reduce costs, and minimize resource waste at each stage. Industry practitioners should strengthen the application of new digital technologies, transition from “industrial thinking” to “digital thinking”, test the applicability of innovative digital technologies in practice, and provide guidance for technological innovation theories.

(4)

Study edge: customer profiling and demand-driven sustainable supply chain management.

The keyword “prediction” has frequently appeared in temporal graphs and trend analysis in recent years. Traditional cross-border agricultural food markets rely on producers and service providers to actively supply goods and services, leading to mismatched supply and demand and resource waste. Therefore, future scholars and industry stakeholders should increasingly focus on customer demands. Firstly, a precise analysis and prediction of consumer demands should be conducted using platform data and advanced data analysis techniques to support the sustainability of the cross-border agricultural food supply chains. Additionally, customer profiling models can provide an in-depth understanding of customers’ purchasing habits, preferences, and changing demands, guiding supply chain strategy adjustments. Lastly, establishing continuous user feedback mechanisms using digital technologies can better meet market demands and enhance the resilience and flexibility of cross-border agriculture. An in-depth exploration of the research frontiers can integrate customer demands with the sustainable development goals.

6.4. Research Challenge in Future

Based on the above research conclusions and analysis, combined with future research directions, this section discusses the specific challenges that may be faced when pursuing these research directions, which need to be addressed in future research.

(1)

Improvement and upgrading of digital infrastructure

Although there has been some progress in digital infrastructure, there are still many problems in the cross-border agricultural supply chain, such as lagging infrastructure construction and inconsistent technical standards in many regions or countries. In the future, with the further development and expansion of the supply chain, the demand for digital infrastructure will continue to increase. Improving and upgrading digital infrastructure and improving its coverage and service quality will be a key challenge for the sustainability of cross-border agricultural product supply chains enabled by digital technology in the future.

(2)

Safeguarding and strengthening data flow security

With the increasing application of digital technology in the supply chain of agricultural products, the security of the data flow is becoming more and more prominent. With the increase in the complexity and scale of the supply chain, the security of the data flow will face greater challenges. How to ensure the security, integrity, and privacy of data in the flow process and prevent data leakage and abuse has become a key research issue for many scholars.

(3)

Response and mitigation of geopolitical risks

The influence of geopolitical factors on the cross-border agricultural supply chain cannot be ignored. In the future, with the changes in the international political and economic situation, geopolitical risks may be further aggravated. Therefore, establishing an effective early warning and response mechanism to reduce the impact of geopolitical factors on the stability and sustainability of supply chains has become an important research topic in this research field.

(4)

Narrowing and bridging the digital divide

The digital divide problem is still prominent in the field of digital technology-enabled cross-border agricultural product supply chains. The differences in the acquisition and use of digital technology in different regions and by different groups may lead to unequal information and unequal resource allocation in the supply chain. In the future, with the globalization of the supply chain and the upgrading of digitalization, the problem of the digital divide will become more prominent. We have reason to believe that narrowing the digital divide will improve the digital level of the supply chain and ensure that all parties can participate and benefit equally.

6.5. Limitations of the Study

Despite our efforts to maintain the scientific rigor and validity of this research, several limitations should be acknowledged. Firstly, the sample literature used for the study was obtained through a thematic search (including titles, abstracts, and keywords), which may have resulted in the omission of relevant articles that did not contain the specified keywords; this could lead to potential sample selection bias and the exclusion of valuable literature. Secondly, the analysis and interpretation of the sample literature were conducted solely using the software “Citespace 6.2.R3”, without comparing and validating the results with other literature analysis software such as VOSviewer 1.6.20 and Histcite v1.1.04. Incorporating multiple analysis tools in the research prospects can provide a more comprehensive and robust analysis. Lastly, although we engaged in discussions and interviews with researchers and practitioners of digital-enabled sustainability, there needed to be more in-depth case studies that could provide insights into real-world practices.