VidSumEval Benchmark Dataset

VidSumEval is a web platform and benchmark dataset for evaluating AI-generated summaries of programming tutorial videos from a learning-centered perspective. The benchmark includes: - 10 curated programming tutorial videos covering C, Java, and Python- 20 AI-generated video summaries produced using VEED and NotebookLM- 50 automatically generated comprehension quiz items- Human evaluation results measuring completeness, clarity, and coherence- Quiz-based learning-effectiveness results from a within-subject study with 20 participants- Prompts, transcripts, benchmark metadata, analysis outputs, and study figures- Source code and reproducibility instructions for the VidSumEval web platform This version updates the artifact packaging, installation documentation, cross-platform setup instructions, licensing information, and platform functionality in response to the ASE 2026 reviews. The dataset accompanies the paper: “VidSumEval: A Web Platform and Benchmark for Evaluating AI-Generated Programming Video Summaries” Paper DOI: 10.1145/3832783.3834594