Peer-Reviewed · Open Access · Open Artifacts
Journal of High School
// Pushing the future of computer science through high-school research
JHCS is a peer-reviewed venue for high-school researchers who ship real computer-science contributions: reproducible ML studies, working systems, verified algorithms, and open-source artifacts reviewed alongside the paper. Two-week turnaround, up to three rounds of mentoring review.
Every JHCS paper ships with a runnable artifact: source code, dataset citations, environment spec, and a script that reproduces the headline result end-to-end. Reviewers don't just read your claim — they clone your repo and run it. That is what “peer-reviewed computer science” means at JHCS.
JHCS is a peer-reviewed journal devoted exclusively to computer science — machine learning and AI, algorithms and theory, systems and networking, security and cryptography, human-computer interaction, programming languages and software engineering, computer vision, NLP, computational biology, robotics, and the adjacent computational fields. Every accepted paper is published with an open-source code artifact and a public dataset citation (or the dataset itself), because in computer science a claim without a runnable artifact is a claim we can't check. Our mission is to identify high-school authors capable of producing reviewable, reproducible CS scholarship, and to hold their work to the same technical bar a university venue would apply.
JHCS publishes evidence-based technical work in one of the four modes below. If a paper does not test a claim with experiments, measurements, or formal analysis — and does not release the code needed to check it — it is outside our scope.
Original ML/AI studies with a well-defined task, honest baselines, ablations, statistical significance, and released training/eval code. Deep learning, RL, NLP, computer vision, safety, evaluation.
Novel systems, tools, or applications with a working implementation, an evaluation against real workloads, and open-source artifacts. OS, networking, databases, distributed systems, embedded, web.
New algorithms or formal analyses with correctness arguments, complexity bounds (upper and lower where relevant), and empirical validation on realistic inputs.
Applied cryptography, systems security, usable-security studies, HCI experiments, computational biology and computational social science — provided the CS contribution is the primary contribution and is evaluated.
Package the paper and code together. A JHCS submission is a PDF and a public GitHub repo that a reviewer can clone, install, and reproduce in under an hour. The repository skeleton looks like this:
Students' time is precious. Round 1 is a decisive go/no-go on novelty, method, and whether the artifact runs. If a paper cannot realistically reach publication, we say so early — with reasons — instead of stringing authors through revisions that won't succeed.
Roughly three in ten submissions advance past Round 1. A pass is a real signal that the contribution is credible and the code is reviewable.
If a paper is not publishable, we say so in Round 1 rather than postponing the inevitable across multiple rounds.
Whether the news is yes or no, you receive a complete, reasoned report within fourteen days — no indefinite waiting.
Volume 1 is open for submissions and publishes on a rolling basis. Placeholder titles below preview the kinds of work JHCS is built to publish; the final table of contents will be updated as papers are accepted. All accepted papers are released alongside their code and dataset artifacts.
Machine learning, algorithms, systems, cybersecurity, human-computer interaction, computer vision, NLP, theoretical computer science, computational biology, robotics — and the adjacent computational sciences. Every submission must present original data, an original artifact, or original analysis, and every one is judged under blinded review.
Transparent, time-boxed, and capped at three rounds. A $100 fee applies to each review round. Reviews weigh algorithmic/systems novelty, empirical rigor, and whether the code artifact actually runs.
An editor confirms scope, format (ACM sigconf / IEEE conference), anonymization, and artifact availability; the round fee is paid in the portal.
A critical go/no-go on novelty of contribution, empirical rigor, and artifact reproducibility. Only about 30% advance; papers that cannot become publishable are declined here.
Reviewers verify that major concerns are resolved, that additional baselines or ablations were run, and that the artifact still installs and executes cleanly.
Final polish, license verification (CC BY 4.0 text, MIT/Apache/BSD code), and reproducibility spot-check — accept, accept with minor edits, or decline.
Every author receives a complete review report within 14 days of each submission or resubmission.
A public GitHub repo with code, dataset citation, environment spec (requirements.txt / environment.yml / Dockerfile) and a README.
A $100 fee per round is paid securely via PayPal from your portal, before each round proceeds.
Open the Portal →PhD, Computer Science
The Editor-in-Chief oversees the review process and the final decision on every manuscript, applying university-level standards for algorithmic novelty, empirical rigor, and artifact reproducibility to a journal built for high-school authors.
The Journal of High School Computer Science exists for one kind of work: research in which a claim is tested by an experiment, a system, an algorithm, or a model — and in which the reader is shown the artifact (the code, the data, the environment) that lets someone else check it. We publish studies in which young researchers pose a clear technical question, build or measure something to answer it, and let the evidence — not marketing prose — carry the argument. We do not publish opinion pieces, un-evaluated prototypes, or write-ups of tutorials someone else already wrote.
This focus is deliberate. Computer science advances because someone releases a benchmark, ships an open-source implementation, publishes the seeds and the hyperparameters, and lets the community reproduce and extend the result. High-school students are fully capable of this craft, but they are rarely given a venue that treats their work as real CS research rather than a science-fair project. JHCS is that venue. Every manuscript is read against the same standards a peer-reviewed CS conference would apply: Is the problem well-posed? Are the baselines fair? Are the ablations honest? Does the code actually run?
We pair that rigor with genuine support. Our reviews are detailed and constructive, returned within two weeks, and designed to teach as much as to judge. A decisive Round 1 means we will not string an author along through revisions that cannot succeed; when a paper can become publishable, our feedback is a map for getting there. The goal is not to gatekeep but to hold the bar high and then help students reach it.
If you are a high-school researcher with a hypothesis, a working notebook, and the discipline to ship a repo alongside your paper, this journal is built for you. We look forward to reading (and running) your work.
[Editor-in-Chief] · Editor-in-Chief
Track editors are PhD-level researchers with peer-reviewed publications in computer science subfields, organized by area: Machine Learning & AI, Algorithms & Theory, Systems & Networking, Security & Cryptography, Human-Computer Interaction, and Programming Languages & Software Engineering. The full roster is available here.
PhD faculty and industry researchers with peer-reviewed publications in ML, deep learning, NLP, computer vision, and reinforcement learning.
Systems researchers and theorists oversee papers on algorithms, distributed systems, databases, programming languages, and networking.
Editors across cryptography, systems security, HCI, computational biology, and computational social science ensure interdisciplinary submissions are held to the same technical bar.