Needlens: An evidence-grounded user-need knowledge graph for Kano-aware design insight and early product-defect warning from large-scale online reviews

This record contains the code and supporting materials for the manuscript “NeedLens: An Evidence-Grounded User-Need Knowledge Graph for Kano-Aware Design Insight and Early Product-Defect Warning from Large-Scale Online Reviews.” It includes the MATLAB implementation of the NeedLens pipeline, configuration files and random seeds, controlled vocabularies and lexicons, induced knowledge graphs, product-family definitions, event-screening and control-matching records, warning-evaluation results, annotation materials, a complete 1,000-sentence adjudicated gold standard, figure-generation scripts, final figures, and verification utilities. Raw Amazon review archives are not redistributed. The provided download script retrieves the required inputs from the public Amazon Reviews 2023 repository and verifies their file sizes and SHA-256 checksums. The files are currently restricted and will be made publicly available when the manuscript enters external peer review.