JHCS publishes peer-reviewed computer-science research by high-school authors. Volume 1 is forthcoming in 2026 and features first-author student papers spanning machine learning, algorithms, systems, cybersecurity, and human-computer interaction. Every accepted paper is published together with its full source code and reproducible experimental artifact.
Volume 1 opens for submissions in 2026 and publishes on a rolling basis. The final table of contents will appear here as papers are accepted; the entries below are [Sample titles pending Volume 1 selection] and are illustrative only. On acceptance, each entry links to the paper PDF and to a public GitHub artifact with code, dataset, and environment spec.
Editor-in-Chief · [TBD]
A brief editorial framing JHCS's mission — why a specialist venue for peer-reviewed high-school CS research, with mandatory open code and artifacts, raises the bar for what a student paper can be. Explains the CS-specific rubric (novelty, empirical rigor, reproducibility, software-engineering quality) and how the review process is designed to teach as well as judge.
Illustrative abstract only. Real abstracts appear when papers are accepted. Sample titles pending Volume 1 selection.
We study how well few-shot classifiers generalize when the source and target datasets differ in feature distribution. We evaluate a set of pretrained tabular foundation models on three public benchmarks under controlled shift, report calibration and error decomposition, run three-seed experiments with 95% bootstrap confidence intervals, and release code, fixed data splits, and an evaluation harness.
Illustrative abstract only.
We present the design and evaluation of a small WebAssembly-based runtime for on-device transformer inference in the browser. We benchmark against WebGPU baselines across three model families, characterize latency and memory footprint under different quantization schemes, and open-source the implementation with a Dockerized build.
Illustrative abstract only.
We give a streaming approximation algorithm for maximum coverage with sublinear memory in the ground-set size, prove a (1−1/e)−ε approximation guarantee under mild assumptions, and validate the bound empirically on a real dataset of browser-tab activity.
Illustrative abstract only.
We survey a stratified sample of high-school students on password practices and cross-reference reported reuse with public breach dumps under strict anonymization, following an IRB-equivalent ethics protocol. We release aggregate statistics, the anonymization pipeline, and the survey instrument.
Illustrative abstract only.
A two-week within-subject study of LLM-assisted note-taking in high-school classes, measuring recall, self-reported cognitive effort, and note-quality against handwritten and template-guided baselines. We report significance-tested effects and release the analysis notebooks.
Illustrative abstract only.
We evaluate cross-domain transfer strategies (fine-tuning, adapters, LoRA) for medical-image segmentation when only a small labelled target set is available. Baselines include a from-scratch UNet and a foundation-model backbone. We report Dice and Hausdorff distance with error bars and release the training pipeline.