1 Introduction

With the profound integration of digital technology across various industries, the digital economy has emerged as a significant factor in the restructuring of the global economic landscape [1]. Countries around the globe have begun to recognize the impact of digital transformation on industries, and more than 80 countries have formulated and implemented industrial policies to promote digitalization [2]. These countries primarily provide support for digital transformation through regulatory laws, technological R&D, and infrastructure construction [3]. At present, the concepts of digital technology and industrial digitalization continue to expand. From a broad perspective, any technology that can drive the digitalization and intelligent operational management of enterprises can be referred to as a digital technology [4]. However, from a narrower perspective, digital technology primarily encompasses artificial intelligence (AI), blockchain, cloud computing, and big data (referred to as the ABCD technologies), as well as mobile communication and the Internet of Things (IoT) [5]. These technologies converge and jointly empower the digitalization of industries. Among them, blockchain is the technology with the most recent origin, fastest development, and most promising application prospects [6]. The world is paying close attention to the development of blockchain technology.

As early as August 2017, the United Nations released a report titled Usage of Blockchain in the UN System: A Desk Review, which expressed its goal of accelerating the Sustainable Development Goals (SDGs) using blockchain [7]. In July 2020, the United Nations published Blockchain Applications in the United Nations System: Towards a State of Readiness, which added details regarding the specific applications of blockchain in the context of digital transformation [8]. In November 2021, the United Nations Conference on Trade and Development released a report titled Harnessing Blockchain for Sustainable Development: Prospects and Challenges. This report detailed the current technical advantages and application challenges of blockchain, and explored the possibilities for an innovative ecosystem based on blockchain [9]. China was also urged to use blockchain to promote the sustainable development of the industrial digital economy. In December 2022, the China Academy of Information and Communications Technology published the Blockchain White Paper. The report analyzed the ecological structure and typical application scenarios of the blockchain industry and looked forward to the future development of blockchain [10].

However, most existing reviews and surveys of blockchain technology are superficial statistical summaries or qualitative analyses, and they fail to truly capture the developmental trajectory of blockchain [11]. Therefore, this study aims to utilize bibliometric analysis to visualize the research hotspots in blockchain. By constructing knowledge maps, we seek to explore the future trends of blockchain development [12]. Overall, the contributions of this study are described as follows.

(1) The study reviews the origin and development of blockchain, presents its concepts, classifications, and main features, and systematically highlights its importance in driving digital transformation in industry. (2) The study uses CiteSpace to analyze the literature in the Web of Science core database and explores knowledge-map visualizations of core information such as the authors and institutions of blockchain-related publications. (3) Combining bibliometric analysis with existing scholarship, the study discusses how blockchain has effectively facilitated digital transformation in active fields of research such as finance and supply chains, and it looks at the prospects of blockchain applications in digital industries such as e-healthcare and the Internet of Vehicles.

The structure of this paper is organized as follows: Section 2 reviews the development and attributes of blockchain and systematically describes the importance in driving the digital transformation of industry. Section 3 focuses on the CiteSpace-based bibliometric methodology and data sources. Section 4 presents the bibliometric results and knowledge maps. Section 5 discusses the digital applications of blockchain in active research fields, and Section 6 presents the findings and implications and discusses the limitations and prospects.

2 Related work

2.1 Review of blockchain development

Blockchain originated in the Bitcoin cryptocurrency, which was first proposed in Satoshi Nakamoto’s paper, “Bitcoin: A Peer-to-Peer Electronic Cash System,” and later applied in electronic money trading systems [13]. A blockchain is asymmetrically encrypted by hash functions, packaged, uploaded to a distributed node, and formed into a chain structure to realize an encrypted record of transaction information, its data maintenance process is illustrated in Fig. 1. Blockchain differs from Bitcoin in that it combines techniques from computer science, cryptography, and mathematics; it includes underlying technologies such as peer-to-peer (P2P) transmission, hashing algorithms, consensus mechanisms, and smart contracts [14]. Blockchain does not rely on third parties for information processing, and it possesses characteristics of decentralization, distributed storage, resistance to tampering, security, and traceability [15]. According to its access mechanism, blockchains can be divided into public blockchains, private blockchains, and consortium blockchains [16]. The pace of blockchain development is rapid, and it has been applied in many industries, such as finance, healthcare, and communication; it has become an engine for promoting the advancement of industry, and has gradually expanded into many fields [17]. Table 1 presents a systemic interpretation of the attributes of blockchain technology.

Fig. 1
figure 1

The data maintenance process of blockchain

Full size image
Table 1 Attributes of blockchain technology
Full size table

2.2 Importance of blockchain in industrial digitalization

In general, blockchain technology provides a new foundational framework and solution for the digital transformation of industries, altering traditional models and operational management processes [18]. The importance of blockchain in industrial digital transformation can be summarized by the following four points.

  1. 1.

    Blockchain has established a decentralized trust mechanism. Blockchain features a distributed ledger mechanism, and its decentralized nature eliminates the need for traditional central authorities. Blockchain uses encryption and consensus algorithms to establish a new trust mechanism [19]. This enables digital processes to operate without intermediaries, reducing the complexity of collaboration and transactions.

  2. 2.

    Blockchain has solved the problem of data security and privacy protection. Blockchain adopts distributed ledger and asymmetric encryption algorithms to ensure that data cannot be modified illegally [20]. In digital transformation, multi-source heterogeneous data are integrated, transmitted, and applied. Blockchain protects the integrity of this information, reducing the risks of data leakage and “51%” attacks [21].

  3. 3.

    The openness, transparency and traceability of the process has been realized. A blockchain records the transaction information and operation process, and all participants can check and verify the authenticity of the data at any time [22]. In digital transformation, this high transparency and traceability can ensure the reliability of enterprises in the fields of supply chain management and product traceability.

  4. 4.

    Blockchain implements the automated processing of smart contracts. A smart contract is a decentralized credit security mechanism that sets pre-defined rules and automatic operation conditions [23]. Once the smart contract has been signed, it will run automatically and cannot be modified. Intelligent contracts enable digital business processes to automate operation and improve efficiency and accuracy.

3 Methodology

3.1 Bibliometric methods

At present, the main scientific drawing tools are CiteSpace, UCINET and NVivo. CiteSpace has the most powerful functions for the visual analysis of publication data [24]. It combines techniques such as network analysis, citation analysis, and text mining to help scholars discover research hotspots, development trends, topic evolution, and collaborative relationships in academic fields. CiteSpace is publication knowledge mapping software developed by Chen Chaomei from Drexel University in the Java language [25]. It can analyze literature authors, core journals, publishing institutions, and citations through coupling, co-citation, and other metrological methods. Moreover, it is capable of leveraging algorithms such as the log-likelihood rate (LLR) for data analysis and text mining [26]. This software uses the visualization of knowledge graphs to outline the evolution of a specific research field within a certain period. Through CiteSpace, scholars can visually analyze a large amount of literature from multiple perspectives, and conduct qualitative analysis based on various knowledge maps in the research field to find the core themes hidden in big data [27]. This study used CiteSpace V6.2.R4 to analyze the hotspots and trends in blockchain research. Figure 2 illustrates the framework design of the analysis process.

Fig. 2
figure 2

The framework design of analysis process

Full size image

3.2 Data sources

To ensure the authenticity of the data, publications in the Web of Science database were selected as the data source for quantitative analysis. In the Web of Science database, the SCIE, SSCI, and AHCI parts of the Core Collection database were used as the source of data for the literature search. The “Year Published” parameter was set to 2015–2023, and “Blockchain” was used as the subject term (the “Topic” parameter) for searching. A sample of 14,889 publications records was obtained (the search was conducted until September 15, 2023), and a plot of the distribution of the number of blockchain papers published by countries from 2015 to 2023 is given in Fig. 3.

Fig. 3
figure 3

Number of blockchain publications by country per year (until September 15, 2023)

Full size image

4 Bibliometric results

The 14,889 original publications obtained from the Web of Science database were imported into CiteSpace V6.2.R4 for visual analysis and knowledge map generation. The “Time Slicing” parameter in CiteSpace was set to January 2015-September 2023; “Years Per Slice” was set to 1; “Node Types” was respectively set to “Author,” “Country,” “Institution,” “Reference,” and “Keyword”; and the “Top N” parameter was set to 50 in the “Selection Criteria.” CiteSpace was then run to perform the visual analysis.

4.1 Visual analysis of the core authors

Using the knowledge map visualization, the authoritative experts and core authors with the most literature in the blockchain field can be visually analyzed. In the “Node Types” parameter of CiteSpace, we selected “Author” to analyze and obtain the core author relationship knowledge map (Fig. 4), which contains 614 nodes and 693 edges. Each circle in Fig. 4 is a node; the larger the font size of a node, the more the number of the author’s publication. The color of the name of the circle indicates the author’s centrality: a darker outline indicates higher centrality. A higher degree of centrality and number of citations indicate the impact of the node within the co-occurrence network is high and it is a focus of attention [28]. The three authors with the highest number of published studies are Neeraj (119 publications), Tanwar (107), and Choo (105). The most prolific authors and their institution along with the total number of blockchain studies for each are listed in Table 2.

Fig. 4
figure 4

Core author relationship co-occurrence map

Full size image
Table 2 Top-10 authors who have published the highest number of blockchain studies
Full size table

4.2 Distribution of blockchain countries and institutions

In CiteSpace V6.2.R4, for node type, we respectively selected “Country” and “Institution,” to investigate blockchain research by country (or region) and institution. The attributes of blockchain literature were analyzed visually by obtaining a knowledge map of the publication collaboration relationships at the national (or regional) level (Fig. 5) and the research institution level (Fig. 6). As shown in Table 3, most of the blockchain literature was published in China (3,903 publications) and the United States (2,091). The three countries with 800–1,500 publications are India (1,452), United Kingdom (914), and Australia (873). The top-10 research institutions include six in China, two in the United States, one in Egypt, and one in Saudi Arabia. The top-5 institutions are the Chinese Academy of Sciences (326), Beijing University of Posts Telecommunications (282), Xidian University (229), King Saud University (223), and the University of Electronic Science and Technology of China (215). This analysis shows that China’s research institutions have made a great contribution to the development of blockchain. The United States, India, and other countries have also published a large amount of research on blockchain. Due to space limitations, the publication statistics for “Country” and “Institutions” in the field of blockchain can be found in the Supplementary Material. Alternatively, you can download all the statistical tables of this paper via https://github.com/ws8228/blockchain.git.

Fig. 5
figure 5

National collaboration relationship co-occurrence map

Full size image
Fig. 6
figure 6

Institutional collaboration relationship co-occurrence map

Full size image
Table 3 Top-10 countries according to the number of published blockchain studies
Full size table

4.3 Visual analysis of highly cited literature

The CiteSpace software can analyze highly cited publications; explore the author feature, citation frequency, and degree centrality of highly cited publications; and automatically cluster the literature research content using the log-likelihood rate algorithm to generate visualization maps of highly cited literature. These help us to visually understand the theoretical basis, evolution, and development directions of related publications [29]. Using CiteSpace and setting the node type to “Reference” (Fig. 7), a total of 653 nodes and 1,397 edges were found. One publication was cited 1,730 times. There were four publications cited more than 1,000 times, and 16 publications were cited 500–1,000 times (the top-10 most highly cited studies are listed in Table 4). The four publications cited more than 1,000 times are analyzed in detail below.

Fig. 7
figure 7

Highly cited publications co-occurrence map

Full size image
Table 4 Top-10 most highly cited studies on blockchain
Full size table

The most highly cited publication is “Blockchains and smart contracts for the Internet of Things,” by Christidis of North Carolina State University [30]. This publication has been cited 1,730 times. The research describes the working principles of blockchain and attempts to apply blockchain to the IoT to promote the sharing of resources and services, solve the application problem of distributed storage under the premise of encryption authentication, and realize the expected value of blockchain in the IoT field.

The second most highly cited publication is the review “Industry 4.0: State of the art and future trends,” by Xu of Old Dominion University [31]. This review states that Bitcoin initially entered the electronic trading market in the form of digital currency, but blockchain is conceptually broader than Bitcoin, and a blockchain can form a decentralized ledger that cannot be illegally modified. It also analyzed the impact of blockchain on industries such as finance, supply chains, and healthcare.

The third most highly cited publication is “Blockchain challenges and opportunities: A survey,” by Zheng of Sun Yat-sen University [32]. This review proposes the use of blockchain to help users manage and store personal private data without having to trust third parties such that companies or individuals are more focused on information applications. In addition, the distributed data platform can protect sensitive information at the root and embed the specification regulations into the information platform in the form of smart contracts to effectively prevent data theft.

The fourth most highly cited publication is “Blockchain technology and its relationships to sustainable supply chain management,” by Saberi of Worcester Polytechnic Institute [33]. The paper points out that emerging smart contracts can be applied not only to the cryptocurrency system, but also enables untrusted participants to securely transmit data or financial transactions without a trusted third party. Furthermore, Saberi explored a new theory of supply chain management method. The proposed supply chain system automatically generates an encryption protocol through a compiler and uses asymmetric encryption to record data transmission information in the blockchain to ensure user transaction privacy.

5 Discussion

5.1 Active blockchain research fields

Keyword analysis is the basis for exploring popular areas of scientific research, and it can also reflect the main ideas and key information in the literature. By analyzing keyword co-occurrence in the blockchain literature, it is possible to explore active research fields and knowledge structures in blockchain research. We selected “Keyword” in CiteSpace to conduct a co-occurrence analysis of the keywords in the literature, employing the LLR algorithm to derive the high-frequency keywords. The default visual layout is the keywords co-occurrence map, and the “Threshold,” “Font Size,” and “Node Size” parameters were set to 10, 5, and 2, respectively. Then we obtained a keyword co-occurrence map (Fig. 8) and timeline map of keyword clusters (Fig. 9). Each circular node in Figs. 8 and 9 represents a keyword. A larger node represents a keyword with a higher frequency of occurrence, and a darker color represents a higher centrality. This visualization can specifically reflect the active topics in blockchain research. The following subsections explore the following four fields: financial transactions, the IoT, supply chains, and mobile communications.

Fig. 8
figure 8

Keywords co-occurrence map

Full size image
Fig. 9
figure 9

Timeline map of keyword clustering

Full size image

5.1.1 Applications in financial transactions

Blockchain is considered to be the basis of Bitcoin. By exploring the potential value of blockchains and distributed ledgers in the capital markets, financial innovation researchers discovered that blockchain encryption protocols and distributed ledger technology have high application value in the banking securities industry [40]. However, from transaction security to information privacy, financial innovation researchers have required additional technological innovations in blockchain [41]. Using the characteristics of blockchain process security and data reliability, central banks and financial institutions could build a new payment system and financial platform (Fig. 10). Citibank N.A. took the lead in researching distributed ledger technology and blockchain smart contracts and found that blockchain has strong potential in the electronic trading and financial securities industry. Because it is a method that records transactions in distributed ledgers, blockchain allows parties to use dynamic asset tracking protocols and multiple nodes to jointly verify and maintain information, thereby simplifying the verification process within and across third-party platforms [42]. Moreover, Industries such as financial securities can use asymmetric ledgers to asymmetrically encrypt transaction information and share data information with partners. In contrast to the traditional process, smart contracts ensure that all financial transactions comply with legal regulations in the form of computer algorithms and automatically enforce the terms of the contract, ensure service efficiency and information transparency, and avoid financial fraud incidents [43].

Fig. 10
figure 10

The financial system based on blockchain

Full size image

5.1.2 Applications in the IoT

The wave of digital transformation has prompted the large-scale application of blockchain in the IoT industry. IoT smart devices must have a pre-built encryption program to prevent the device from being attacked when it is in standby mode [44]. To address the security of embedded devices in IoT, operators can use the blockchain to upgrade some firmware systems. The advantage of this scheme is that it enables repairs to firmware vulnerabilities in a timely and effective manner without the need to upgrade the entire network to prevent hackers from attacking the IoT [45]. IBM Corporation used blockchain when building the Watson IoT professional service platform to manage smart devices; the platform converts the device data into a format that can be recognized by the application’s programming through hash functions, and it satisfies smart contracts (see Fig. 11). The information of a smart contract is sent to the blockchain ledger, and the devices in the IoT can share information in real time [46]. Filament Technology, an IoT solution provider, has launched a wireless sensor called Taps that is based on blockchain. The communication module of the Taps device is technically guaranteed by the blockchain so that it can be wirelessly connected to smart devices such as cell phones and near field communication within a specific range [47]. It has a broad application market in digital farms and AI factories.

Fig. 11
figure 11

The IoT platform based on blockchain

Full size image

5.1.3 Applications in the supply chains

In the process of product supply chain management, there are many participants, such as suppliers, retailers, and logistics providers. It is relatively difficult to realize the traceability of products in real time [48]. Therefore, the information traceability feature of the blockchain can be applied to the supply chain management of goods, which makes it convenient to track and request the relevant information about the goods during production and distribution [49]. To eliminate the risk of illegal modification of the product information, the provider of the traceability service provides product traceability services as an independent third-party, opening query function of product transmission and storage to buyers via a P2P network. Su and Wang [50] explored the use of blockchain to match the digital identification of agricultural and livestock products, enabling consumers to query the origin and production date of goods through the blockchain. This allows for the traceability of the entire process of logistics and sales of commodities. Figure 12 depicts the supply chains management process based on blockchain. The purpose of tracing the origin of agricultural and livestock products and sharing transaction records is to help consumers combat commercial fraud. Because digital transformation is widely used in farm management, blockchain promotes the information traceability and data sharing of agricultural and livestock products, and this technology provides a reliable third-party platform for supply chain managers [51]. The supply chain ledger automatically audits product information with the aid of a smart contract, which improves the transparency and credibility of the entire traceability process and reduces the maintenance cost of the supply chain information.

Fig. 12
figure 12

The management process of blockchain-based supply chains

Full size image

5.1.4 Applications in mobile communications

Secure and effective encryption to protect communication information has become the primary problem facing the 6G industry [52]. To date, communication operators have studied the data encryption and transmission characteristics of blockchain and attempted to apply this technology in 6G communications to solve problems such as data packet loss, network delay, and server congestion in mobile communications. The confidentiality and traceability features of the blockchain ledger ensure the privacy of the user’s communication records to achieve mutual trust between the information provider and recipient [53]. Verizon Communications has begun to combine asymmetric encryption technology and access control policies to reduce the complexity of key management and ensure that user communication information is secure. However, blockchain applications in communication networks are affected by the “Proof of Work” problem, which increases the demand for computing power and storage space in mobile devices [54]. To solve this problem, a mobile edge computing (MEC) device based on the blockchain framework was designed by Bell Laboratories. The computationally intensive MEC mining task can link to the nearest edge computing node to create an encryption zone [55]. The block is cached in the MEC server at the fastest speed to protect the privacy of the communication information. Figure 13 shows the model of blockchain application in 6G mobile communication.

Fig. 13
figure 13

The model of blockchain application in 6G

Full size image

5.2 Development prospects of blockchain

Bursting keywords can indicate the research topics whose related disciplines have received a high level of attention at a certain time. These disciplines can produce important scientific research value or can make major breakthroughs in future scientific research projects. CiteSpace excels in detecting bursts in research fields, providing valuable insights into significant breakthroughs [56]. After sorting the weights of bursting keywords, we constructed a co-occurrence network and timeline graph, and then analyzed the frontier areas of a certain subject and conducted in-depth research. Based on this, we selected “Keyword” as the node type in the CiteSpace software, derived the top 15 bursting keywords in blockchain literature through the “Burstness” option (see Fig. 14). Moreover, we set the “Threshold,” “Font Size,” and “Node Size” parameters to 15, 5 and, 30 respectively, and performed visual analysis on the bursting keywords, and selected the “Time-zone View” option to obtain a time zone map of the blockchain highlights (see Fig. 15). Figure 15 shows that the development prospects of blockchain research in the decade can be divided into three stages: 2015–2017, 2018–2020, and 2021–2023.

Fig. 14
figure 14

Top-15 strongest keywords in the blockchain literature

Full size image
Fig. 15
figure 15

Time zone map of bursting keywords

Full size image

In particular, during 2015–2017, increasing numbers of researchers studied the characteristics of blockchain and laid the theoretical foundation for this technology. During 2018–2020, research on blockchain by multinational researchers showed an explosive increase. Scientists from around the world seemed to believe that blockchain would become a basic technology in the industrial digital transformation, and devoted their energy to the blockchain application, conducting research and applying blockchain to various fields such as financial currency and IoT. During 2021–2023, the range of applications of blockchain gradually incorporated the field of AI and digital transformation. The following sections focus on blockchain in the fields of smart grids, e-healthcare, Internet of Vehicles, and machine learning.

5.2.1 Developments in smart grids

Blockchain is expected to facilitate demand analysis for the development of the energy Internet, the ubiquitous power IoT and the carbon emissions trading market. Moreover, it is anticipated to provide technical support for the distributed management of novel power systems such as virtual power plants, smart micro-grids, and energy storage stations [57]. The Duke Energy Corporation is attempting to apply blockchains to their power systems and establish a machine-to-machine electricity market. Participants in this market include power plants, power grid companies, and power consumers. Power plants compile and count comprehensive power generation parameters, calculate and announce the quotation’s price of electrical energy transactions. After obtaining a quote, the grid company conducts a price analysis and meets the energy demand of the electricity consumers at the lowest cost [58]. General Electric Company uses blockchain to analyze the consumption features of electricity consumers to provide them with professional energy services. They also began to use smart contracts to predict users’ energy consumption and store the data in distributed ledgers [59]. To protect data privacy in smart grids, the State Grid Corporation of China proposed a blockchain-based privacy protection and power data aggregation scheme using a bloom-filter data-deduplication algorithm to achieve rapid authentication of blockchain nodes. In contrast to traditional authentication schemes, the data aggregation of blockchain-based smart grids system is much more computationally efficient (see Fig. 16). It also improves the identity verification algorithm for system initialization and shortens the authentication time [60].

Fig. 16
figure 16

The model of blockchain-based smart grids system

Full size image

5.2.2 Developments in e-healthcare

With the increase of medical diagnosis data and the development of remote patient monitoring systems, the security of medical healthcare data has attracted widespread attention in the medical community [61]. This community is attempting to use blockchain to protect patient medical information and establish a blockchain-based medical data preservation system (MDPS), the model of MDPS is shown in Fig. 17. The MDPS relies on the data storage and encryption algorithm of a blockchain to store important medical data, and verify the authenticity of the original data [62]. MDPS can process the protected health information (PHI) of a patient in a timely manner and close security loopholes in remote monitoring. Abbott Laboratories has begun building a MDPS using a private blockchain and to enhance data analysis and the management of medical sensor security. With the support of a blockchain, smart sensors and health equipment can provide doctors with patient health data in a timely manner through smart contracts while also performing real-time monitoring and medical guidance [63]. Sharing e-health records can improve the accuracy of medical diagnosis. To balance e-health record sharing and personal health privacy, the PHI assistant system based on the consortium blockchain is gradually promoted in the healthcare system, thereby realizing the privacy security of PHI data [64].

Fig. 17
figure 17

The model of MDPS based on blockchain

Full size image

5.2.3 Developments in the internet of vehicles

The Internet of Vehicles is an extension of smart electric vehicles, IoT and AI. It provides new services that benefit the driver and car suppliers; however, intelligent e-vehicles and Vehicle-to-Everything (V2X) communications are also threatened by security and privacy risks [65]. The use of blockchain asymmetric encryption and smart contract technology is expected to solve potential problems such as the illegal tracking and remote hijacking of vehicles while enhancing the security of driverless and V2X communication frameworks. The prerequisite for realizing autonomous driving and smart transportation is to protect user privacy [66]. Jidu Auto Company proposed the Credit-Coin model, which is based on blockchain and publishes or transmits real vehicle traffic information through the V2X system. The Credit-Coin model enables user car owners to send anonymous notifications, but it will report false information and warn the publisher. Moreover, users who publish effective data will receive point-based rewards. This trust mechanism-based model based will inspire users to share vehicle and traffic information in untrustworthy environments [67]. In addition, the intelligent transportation of electric vehicles involves real-time interaction problems in car owner information, urban transportation, and charging facilities. Using blockchain in the electric vehicle cloud and edge computing can mitigate various vulnerabilities of e-vehicles such as the hijacking of multi-source heterogeneous data transmission and communication systems, and effectively ensure the driver privacy and data interaction in this type of computing (see Fig. 18). Trust, privacy, and security are important factors influencing V2X communication protocols, and these factors mean that the prospects for blockchain-based applications via the Internet of Vehicles is strong [68].

Fig. 18
figure 18

The framework of blockchain-based V2X system

Full size image

5.2.4 Developments in machine learning

Since 2021, which marked a surge in the popularity of multilingual pre-trained language models such as Chat-GPT, AI machine learning has once again gained public attention. As a bursting field of convergence, the application of blockchain in machine learning is currently in a preliminary exploration stage [69]. Because it combines existing theoretical achievements, blockchain has the potential for the following three applications in machine learning. Firstly, for deep learning, blockchain can provide data storage and preservation functions for decentralized and non-illegal modifications, effectively ensuring the privacy and security of sensitive data during the transmission process in artificial neural networks [70]. Secondly, for natural language processing (NLP) models, blockchain-based smart contracts can automatically identify the semantics of the text and complete the processing and understanding of the text. It can automatically execute NLP programs such as text classification, machine translation, and sentiment analysis according to predetermined procedures, and will not be affected by external interventions [71]. Thirdly, for public opinion monitoring and prediction, the blockchain P2P data transmission method includes a hash value and timestamps in the data of each block. This can effectively verify whether public opinion information is true in order to accurately predict the future direction of public opinion [72]. The applications of blockchain in AI machine learning are still in the experimental accumulation stage (the framework is shown in Fig. 19), but there is no doubt that the combination of these topics will help accelerate the digital transformation of the industry.

Fig. 19
figure 19

The framework of blockchain application in machine learning

Full size image

6 Conclusions

6.1 Findings

We used the Web of Science core collection with “Blockchain” as the keyword for the literature search and CiteSpace for data analysis, and explored the active research fields and future evolving trends of the blockchain literature; the following are the findings of this study. (1) We systematically explained the essence, concepts, features, main classifications, and core technologies of blockchain and analyzed the important role that blockchain plays in facilitating the digital transformation of industry. (2) We visualized the knowledge maps of key information such as author, institution, and highly cited literature in blockchain-related publications, and described the characteristics of the bibliometric results. (3) We performed an analysis of the keywords in the literature and discussed in detail the current application status of blockchain in the industries of financial transactions, the IoT, supply chains, and mobile communications, as well as the prospects for development in the fields of smart grids, e-healthcare, Internet of Vehicles, and machine learning. Next, the paper extensively explores four implications brought by the development of blockchain.

6.2 Implications

Promoting the digital transformation of blockchain. Digital technologies such as blockchain, AI, and machine learning have become the main engines that are driving the development of Industry 4.0. These core technologies should be used to perform all-round digital transformations of traditional industries, and promote the integrated development of the digital industry with the real economy. Enterprises should seize the commanding strategic heights of digital transformation, vigorously introduce the relevant technical talents, improve their R&D capabilities for key digital systems and software, and open up major channels for SDGs and circular economy.

Encouraging technological innovation of blockchain. In addition to the relatively mature applications of blockchain in the field of electronic currency trading, other industry applications are only at the stage of trial operation or concept application. However, the theory of this study shows that the distributed storage, confidentiality, and easy traceability of blockchain can be applied in many fields and can promote industrial innovation. Therefore, it is recommended that the government support market guidance and encourage blockchain development and business. Innovation drives continuous improvement in blockchain, strengthening its relationship with physical industries and enabling practical implementation across various applications.

Advancing the depth of blockchain research. The academic community has initially explored the characteristics and advantages of technology, and has achieved initial results in digital transformation of the supply chains, mobile communications and other fields. However, the question of how this technology can apply its disruptive value to the digital transformation of enterprises requires further exploration. In addition, the questions of whether or not blockchain can achieve fully decentralized applications, and whether or not the existing node computing power can meet the P2P transmission requirements of the entire network require further study, so it will be imperative to expedite research on blockchain digital applications.

Strengthening the supervision of blockchain industry. Blockchain development has short development cycles and complex technological structures, yet there are still some limitations at the practical application level. Hence, it is crucial to foster innovative development of blockchain digitization while ensuring manageable risks and implementing necessary regulatory oversight. Government departments, regulatory agencies, and industry associations should collaborate to establish and enhance industry standards and development strategies for innovative blockchain applications. This collaborative effort ensures the healthy and sustainable development of the blockchain digital industry.

6.3 Limitations and prospects

At present, blockchain research is in an exploratory period, and its related applications are mixed on the market. Although blockchain research has achieved preliminary theoretical results, the academic community has not established a perfect unified technical standard. Therefore, blockchain should not be limited to Bitcoin mining and speculation; instead, the existing technological achievements should be used to solve real-world problems such as industrial digital transformation and circular economy development. What is certain is that blockchains have increasingly gained attention and recognition by industry. With continuous exploration by the academic community, it will realize its great potential in the future. We will continue to pay attention to the latest research results; develop the practical applications of blockchain in V2X, AI, machine learning, and convolutional neural networks; and accelerate the process of SDGs and industrial digital transformation.