Scientometric Analysis Unveils Rapid Growth in Quantum Machine Learning and Cybersecurity

A comprehensive scientometric evaluation examining the intersection of quantum computing, cybersecurity, and machine learning has revealed rapid publication growth and structural consolidation across the multidisciplinary domain. Published in Discover Networks by researchers Basil Hanafi and Mohammad Ali, the study analyzes 495 peer-reviewed articles indexed in Scopus and Web of Science from 2018 through mid-2025. The investigation maps how quantum computation, algorithmic learning, and cryptographic defenses converge to address challenges in secure communication, complex optimization, and probabilistic decision systems. Although earlier scholarship treated these technologies within isolated subfields, contemporary literature indicates that algorithmic advances in machine learning, post-quantum defenses, and physical quantum architectures increasingly depend on one another. The analytical framework combines quantitative performance analysis with structural science mapping to evaluate publication volumes, source concentrations, author productivity distributions, institutional involvement, national collaborative partnerships, and conceptual thematic trajectories.

To construct a reproducible evidence base, the investigators executed a structured retrieval workflow guided by PRISMA standards across two major citation databases. The search query targeted indexed peer-reviewed journal articles and conference proceedings published in English between 2018 and mid-2025 using coordinated terminology matching quantum computing, security, and machine learning across titles, abstracts, and author keywords. The initial search yielded 583 records, comprising 453 entries from Scopus and 130 from Web of Science. A two-stage deduplication procedure was applied to clean the collected metadata. In the first phase, five duplicate or redundant records were eliminated during preliminary preparation before merging outputs. Following the formal merger, bibliographic matching across digital object identifiers, titles, and author details eliminated another 83 cross-database duplicate entries. Scope verification confirmed that all 495 retained publications directly or thematically addressed the triadic intersection, establishing the final analytical dataset.

Temporal performance metrics demonstrate an explosive acceleration in scientific production throughout the analyzed period, transitioning from sporadic exploratory papers to substantial annual publication volumes. In 2018 and 2019, scholarly output remained modest, recording five and four papers respectively, before doubling to 10 publications in 2020. The research field subsequently entered an expansive phase, producing 31 publications in 2021 and 35 in 2022, followed by a dramatic surge to 83 articles in 2023. Annual output reached 165 publications in 2024, representing an overall annual growth rate of 64.36 percent across the survey timeframe. Records for 2025 registered 162 publications within only the first six months, demonstrating sustained publication intensity despite representing an incomplete annual reporting window. This trajectory has yielded an exceptionally young body of literature with an average document age of only 1.35 years, illustrating expanding scholarly engagement across academia.

Analysis of citation patterns reveals distinct temporal dynamics influenced by rapid publication expansion and standard citation latency. The overall dataset exhibits an average citation rate of 6.598 citations per document across all 495 publications. Longitudinal evaluation demonstrates that citation impact was heavily concentrated among foundational contributions published during the earliest years of the analyzed window, reaching a distinct peak in 2021. Articles published in 2018 achieved an average of 17.2 citations, which escalated to 28.1 citations per publication for papers appearing in 2021, alongside an annual citation rate increasing from 2.15 to 5.62 citations per year. Subsequently, calculated citation averages decreased to 18.6 citations per article in 2022, 7.5 in 2023, 4.4 in 2024, and 1.9 in mid-2025. The authors emphasize that this downward curve reflects classical citation lag effects alongside rapid growth in recently published articles rather than declining research relevance.

An examination of global citation influence highlights enduring theoretical benchmarks alongside rapidly diffusing recent contributions within the quantum security literature. The most cited publication in the dataset is the 2022 contribution by Caro and colleagues in Nature Communications, which accumulated 215 citations, an annual rate of 53.8 citations, and a normalized impact score of 11.58. A 2024 paper authored by Ajani in the International Journal of Intelligent Systems and Applied Engineering ranks second with 162 citations, exhibiting the highest citation velocity at 81 citations annually and a normalized score of 37.28. A 2023 article by Ur in Future Internet gathered 121 citations, averaging 40.3 citations per year. Earlier foundational investigations include work by Du published in 2021 in Physical Review Research with 85 citations, followed by 2021 studies by Zhou, Cincio, and Magann gathering between 78 and 82 citations each, alongside high-impact contributions by Bayerstadler, Li, and Duong.

Dissemination channels reflect early consolidation around a concentrated core of publication venues alongside broad representation across computer science outlets. The 495 evaluated documents appeared across 336 distinct journals and conference proceedings. The primary venue by volume is IEEE Access, which contributed 18 articles, followed by Lecture Notes in Networks and Systems with 14 papers, and Communications in Computer and Information Science with 10 publications. Conference sources, notably IEEE Quantum Week / QCE and ACM International Conference Proceedings Series, also represent prominent dissemination channels. Beyond volume, source impact assessment using local h-index and g-index metrics identifies IEEE Access as the leading outlet for citation influence, with secondary impact clusters forming in Communications in Computer and Information Science, IEEE Transactions on Consumer Electronics, and Lecture Notes in Networks and Systems.

The demographic structure of authorship illustrates high researcher participation characterized by significant productivity skewness. Across the 495 examined publications, research was produced by 2,385 individual authors, generating 1,458 distinct author keywords and 2,248 Keywords Plus terms. Demonstrating the collaborative nature of quantum computational research, the dataset contains only 46 sole-authored papers, resulting in an average collaboration density of 5.53 authors per document. However, the international co-authorship rate remains relatively low at 10.3 percent. Author productivity closely matches Lotka’s law, wherein the overwhelming majority of scholars contribute only once to the field, while a limited hierarchy of recurring investigators drives sustained output. Authors NA N and Zhang S emerge as the most prolific contributors within the database. When evaluating local citation impact, researchers Kumar S and Rahman M demonstrate the highest h-index values, while Rahman M, Kumar P, and Pal A display superior g-index metrics.

Institutional participation and national geographic production patterns reveal substantial concentration anchored by major national laboratories and university consortia. Additional institutional contributors include Kennesaw State University, Central South University, and the Chinese Academy of Sciences. National output displays a distinct concentration across the United States, India, and China, which collectively serve as the primary producing hubs and dominant centers for corresponding authors. However, corresponding author records from these leading nations consist predominantly of single-country publications, indicating domestic clustering.

Science mapping techniques, including keyword co-occurrence analysis, co-word networks, and treemap visualizations, demonstrate a stable conceptual architecture organized around central computational and security terminology. The primary intellectual nucleus is defined by quantum computing and machine learning, which exhibit the highest frequencies across the dataset. A secondary tier of recurring concepts includes quantum computers, network security, and quantum electronics, reflecting an essential focus on physical hardware constraints alongside algorithmic models. Keyword co-occurrence networks indicate that quantum computation bridges theoretical computing with cybersecurity, post-quantum cryptography, and adversarial machine learning. In author-keyword networks, quantum computing forms strong ties with quantum machine learning, quantum optics, and quantum cryptography, while linking outward to emergent areas such as healthcare analytics, optimization, sensors, and natural language processing. Longitudinal thematic evolution tracking reveals that previously disparate research streams have converged toward unified quantum computing frameworks, while blockchain maintains a distinct parallel trajectory.

Factorial analysis and thematic mapping categorize domain concepts into core foundational themes, transversal methods, and emerging specialized niches. Foundational themes situated at the central, highly cohesive core include quantum hardware components, quantum channels, qubits, quantum circuits, and quantum electronics, which together provide the physical baseline for the discipline. Transversal methods, such as support vector machines, intrusion detection, and network security protocols, serve as connective analytical infrastructure linking classical defense mechanisms with quantum applications. Meanwhile, subfields including federated learning, differential privacy, post-quantum cryptography, and distributed quantum intelligence emerge as specialized themes with relatively low current publication density. Thematic trajectory analysis also highlights that while broad foundational terms like computation, quantum theory, and classical algorithms characterized early output, later publications favor specific applied terminology including quantum machine learning, quantum electronics, and quantum cryptography, reflecting an advancing progression toward implementation frameworks.

Synthesizing these scientometric observations, the authors articulate six strategic research directions alongside crucial methodological boundaries governing the study. Future research priorities require developing scalable hybrid quantum-classical architectures that optimize computation within noisy intermediate-scale quantum constraints, establishing operational post-quantum cryptographic protocols with rigorous benchmark comparisons, engineering noise-tolerant quantum machine learning models, deploying privacy-preserving federated quantum intelligence across sensitive sectors, and broadening international collaborative networks. Methodologically, the study boundaries must be considered when evaluating these findings. The underlying dataset was restricted to English-language peer-reviewed literature indexed in Scopus and Web of Science, omitting non-English databases such as CNKI, SciELO, and J-Stage, as well as preprint servers like arXiv, HAL, and SSRN. Furthermore, reliance on a fixed phrase query without secondary sensitivity analysis, inevitable citation lag suppressing recent impact metrics, and subjective keyword standardization choices represent acknowledged constraints that define the scope of this baseline bibliometric investigation.

Subject of Research: Scientometric analysis of research at the intersection of quantum computing, machine learning, and cybersecurity

Article Title: Research evolution in security involved for quantum computing and machine learning through scientometric review analysis

Article References: Hanafi, B., & Ali, M. (2026). Research evolution in security involved for quantum computing and machine learning through scientometric review analysis. Discover Networks, 2(1), Article 15. https://doi.org/10.1007/s44354-026-00029-y

Image Credits: AI Generated

DOI: 10.1007/s44354-026-00029-y

Keywords: Quantum Computing, Machine Learning, Cybersecurity, Scientometrics, Quantum Machine Learning, Quantum Cryptography, Bibliometric Analysis, Network Security, Research, evolution, security, involved