About Big Data Mining and Analytics

Big Data Mining and Analytics

Big Data Mining and Analytics is an English-language, open-access journal of data science published by Tsinghua University Press in Beijing. Launched in 2018 and sponsored by Tsinghua University, one of China's leading research universities, it is published quarterly and covers big data infrastructure, data mining, analytics, visualisation and applications of artificial intelligence. It publishes research and reviews from Chinese and international big-data researchers, is distributed under open-access terms with publishing costs subsidised by its publisher so that authors do not pay an article processing charge, and is indexed in the Emerging Sources Citation Index, EI and Scopus. Its print ISSN is 2096-0654 and its online ISSN is 2097-406X.

The journal was founded at a moment when big data had moved from a slogan to an engineering and scientific discipline in its own right. Universities, internet companies, governments and laboratories were all generating and collecting data at scales that older statistical and database methods could not handle comfortably, and they needed methods to store it, clean it, mine it for patterns, analyse it in near real time and present the results intelligibly. A journal that treated that whole pipeline — infrastructure at one end, applications at the other — as a single field was the publishing response to that need, and Tsinghua, with its strength in computer science and its own university press and SciOpen publishing platform, was a natural home for it in China.

Origins, sponsorship and the Tsinghua base

Big Data Mining and Analytics is described by its publisher as an academic journal sponsored by Tsinghua University and published quarterly by Tsinghua University Press. The first issues appeared in 2018, the year from which its Scopus coverage also runs. Tsinghua University Press is one of China's major university presses and the operator of the SciOpen platform, through which the journal is presented and read online alongside other Tsinghua journals such as Tsinghua Science and Technology. Sponsorship by the university itself, rather than by a department alone, signals that the title was intended as a flagship for an interdisciplinary area in which Tsinghua has researchers across computer science, automation, statistics and many application domains.

Beijing is the journal's home in the directory record, and that location is substantive, not merely administrative. The capital concentrates national laboratories, the headquarters and research centres of major technology companies, government data programmes and several of China's strongest universities. A big-data journal edited in that environment sits close to both the producers of very large datasets and the researchers building methods to exploit them, and its contents — which range from systems papers to work on healthcare, manufacturing and the Internet of Things — reflect that proximity to practice as well as to theory.

Publishing model: subsidised open access

The journal's open-access model is one of its most distinctive features. Its submission guidelines state plainly that it is a subsidised open access journal where Tsinghua University Press pays for the publishing costs incurred by the journal, and that authors do not have to pay any article processing charge or open access publication fee. Articles are distributed under the terms of the Creative Commons Attribution 4.0 International License, which allows reuse with attribution. In current terminology this is often called diamond open access: free for readers and free for authors, with costs carried by the sponsoring institution.

That model has practical consequences. Authors from institutions without publication funds face no fee barrier; readers anywhere can read the full text; and the journal can compete for good papers on quality and visibility rather than on an author's ability to pay. At the same time, a subsidised model depends on the sponsor's continuing commitment, which is why the backing of Tsinghua University and its press — institutions with a long-term interest in the visibility of Chinese data-science research — is an important part of the journal's identity and stability.

Scope: from infrastructure to applications

The journal aims, in its publisher's words, at presenting up-to-date scientific achievements with high creativity and great significance in big data mining and analytics, and it welcomes contributions from all over the world. Its scope, as recorded in its directory description, spans big data infrastructure, data mining, analytics, visualisation and applications of artificial intelligence. Scopus classifies it across artificial intelligence, computer networks and communications, computer science applications and information systems, and in those categories it has been ranked in the top quartile, Q1 — a classification that captures its position between core computer science and applied, systems-oriented research.

Infrastructure work addresses how large-scale data is stored, managed, moved and processed: distributed and parallel architectures, databases and data systems, stream processing and the platforms on which mining and learning actually run. Mining and analytics work addresses what is done with the data once those systems exist: pattern discovery, predictive modelling, statistical learning for very large datasets, and methods that remain reliable when data is heterogeneous, dynamic or distributed. Visualisation, the third leg named in its scope, addresses the human end of the pipeline — how analysts and decision-makers explore results and understand what a model or a dataset is showing them.

Data mining, statistical analytics and machine learning

Data mining is in the journal's title, and it remains the intellectual core around which the other topics gather. Classical data-mining questions — classification, clustering, association, anomaly detection and knowledge discovery — reappear at big-data scale with new difficulties: algorithms must be scalable, evaluation must be honest on massive and messy data, and results must be reproducible. The journal's special issues illustrate how it organises that territory. One announced special issue, for example, was devoted to recent advances in statistical analytics theories and methods for big data, covering statistical machine learning algorithms for big data analytics among its topics, with guest editors from Shenzhen University, a Guangdong laboratory for artificial intelligence and the digital economy, and Qatar University — a team that itself shows the journal's Chinese base and international reach operating together.

Other special-issue themes visible in the journal's record include intelligent systems and the Internet of Things, and big data for artificial-intelligence-powered Internet of Healthcare Things, exploring data science, emerging trends and applications in health. Those themes are instructive: the journal does not treat artificial intelligence as a separate silo from big data, but as one of its principal consumers and drivers. Large datasets make modern machine learning possible; machine learning, in turn, is now one of the main tools for mining those datasets. Papers that work at that junction — a learning method designed for data at scale, or a data system designed for learning — are squarely in the journal's territory.

Applications across science, industry and society

Applications are explicitly part of the scope, and the journal's published contents show the range that follows from that choice. Work appearing in its pages has addressed, among many other subjects, recommendation and sparse matrix completion, prediction of solar radiation components using deep learning models trained on meteorological data, and serverless computing costs for federated learning aggregation, studied through a major cloud platform as a case. That spread — media and commerce, energy and environment, cloud systems and privacy-conscious learning — is typical of big-data research, where the same underlying methods migrate quickly between domains.

For readers, the applications strand is what makes the journal useful beyond the data-science specialist community. An engineer working on intelligent manufacturing, a researcher building healthcare data systems, or an analyst studying urban or energy data can find methods presented in a form tested on a real problem, with real data constraints, rather than only on a benchmark. The journal's own calls for its special issues repeatedly invite both original research and review articles from researchers and practitioners, a pairing that reflects an intended audience spanning universities, laboratories and industry.

Peer review and how authors publish there

Manuscripts are selected for publication according to editorial assessment of their suitability and evaluation by independent reviewers; papers are usually sent to two or more reviewers, according to the journal's submission guidelines. Submissions are made electronically through the ScholarOne Manuscripts system, editorial staff edit accepted papers to improve accuracy and clarity, and it is a condition of publication that a manuscript has not been published and is not simultaneously submitted or published elsewhere. Abstracts of about 100 to 150 words are required to summarise a paper's aims, methods, results and conclusions. These are conventional arrangements for a serious international journal, and stating them matters for a directory profile: they are the visible machinery that separates a peer-reviewed journal from a repository or a magazine.

The journal is published quarterly — four issues a year — and its articles carry digital object identifiers in a form that embeds its initials, with the DOI prefix pattern used by Tsinghua University Press on SciOpen. Online-first and just-accepted presentation means new papers can appear ahead of their issue assignment, which for a fast-moving field shortens the time between acceptance and visibility while the quarterly issue structure preserves a curated, browsable record.

Indexing, metrics and discoverability

Big Data Mining and Analytics is indexed in the Emerging Sources Citation Index, part of the Web of Science, in EI, and in Scopus, and it is also indexed by IEEE Xplore and listed in the Directory of Open Access Journals. That combination is unusually informative. ESCI indexing gives the journal a Web of Science presence and, in recent Journal Citation Reports, an impact factor reported in journal-ranking listings in the region of six — a high figure for a journal founded only in 2018, and one that varies by year as all such figures do. EI indexing places it in the engineering literature, IEEE Xplore places its contents in front of the electrical and computing engineering community, and Scopus Q1 rankings in multiple computer-science categories place it in the top quartile of its fields.

Discoverability, for a young open-access journal, is not a bureaucratic detail; it is the mechanism by which the founding purpose is achieved. A paper that can be found in Scopus, Web of Science, EI and IEEE Xplore, and read in full without charge, has a realistic chance of being used by researchers who have never heard of the journal itself. The journal's metrics should therefore be read alongside its access model: subsidised open access plus broad indexing is a deliberate strategy for maximising the international circulation of work in a field where China is a major producer.

Place among Chinese data-science journals

China's English-language computing journals include long-established titles such as the Journal of Computer Science and Technology, founded in the 1980s, and broad artificial intelligence titles such as CAAI Transactions on Intelligence Technology. Big Data Mining and Analytics, founded three decades after the oldest of those, belongs to a newer generation: born open access, born on a modern platform, and focused on a research area defined by scale of data rather than by a traditional department. Its youth is therefore not a weakness in its profile but a marker of its generation — it did not have to convert from print or from subscriptions, because the scholarly-communication world it entered in 2018 already expected both.

Within Tsinghua University Press's own list, the journal complements more general science and technology titles by giving the press a specialist flagship in one of the most cited and most applied areas of contemporary computing. For the wider Chinese research system, it provides a domestic English-language venue of international standing, so that Chinese big-data researchers can publish at home without sacrificing visibility abroad, and international researchers can reach a Chinese readership through a journal that Chinese institutions themselves read and cite.

Readership and summary

The journal's readership is the global big-data research community: computer scientists and statisticians working on mining and learning at scale, systems researchers building data infrastructure, and domain scientists and engineers applying those methods in healthcare, energy, manufacturing, the Internet of Things and other data-rich fields. Its reviews serve newcomers and adjacent specialists; its research papers serve peers working on the same problems; its special issues serve as periodic maps of where a subfield is moving.

In summary, the verifiable profile is this: Big Data Mining and Analytics; Beijing; sponsored by Tsinghua University and published by Tsinghua University Press; launched in 2018; quarterly; subsidised open access under a Creative Commons Attribution licence with no author processing charge; print ISSN 2096-0654 and online ISSN 2097-406X; indexed in ESCI, EI and Scopus, as well as IEEE Xplore and DOAJ; covering big data infrastructure, data mining, analytics, visualisation and artificial intelligence applications. It is a young journal, but one whose indexing, rankings and publishing model make it one of the most visible China-based venues in data science.