SADPTI Dataset: Indonesian Higher Education Domain-Enumeration Metrics (22 Phase-1 records from 51 operational HE web domains; analytic n1=21; lexicon v1.0.0; Phase-B artefacts)

Overview. This dataset accompanies the SADPTI (Scraper Artefak Digital Pendidikan Tinggi Indonesia) manuscript on an auditable enumeration-first framework for Indonesian higher-education digital-artifact acquisition. It releases Phase-1 funnel metrics, purposive case-study CSV archives, a curated academic subdomain lexicon, and Phase-B experimental artefacts under a full-transparency policy for publicly observable HE web infrastructure. Scale consistency. Operational population N = 51 Indonesian HE web domains (47 .ac.id + 4 other institutional TLDs); national corpus 19,854 digital-article candidates (snapshot 14 July 2026). Deposited Phase-1 records: 22 target domains. Analytic sample n1 = 21 independent institutions (Campus6 / home.amikom.ac.id excluded as a pseudo-replicate of Campus4). Full-text page bodies are not deposited. Contents. • campus_id_mapping.csv — Explicit campus_id ↔ public domain mapping (Campus1–Campus22) with analytic_sample flag. • phase1_stats_national_anonymized.csv — 22-record audit export (legacy filename; not an anonymity claim). • phase1_stats_national_anonymized_n21.csv — Analytic subset for national tables. • campus_articles_matched.csv — Valid-host vs article pairs (n = 20). • correlation_pairs.csv — Pearson/Spearman pairs. • case_studies/Campus1|2|3/ — Seven funnel CSVs per purposive case. • fase_b1_arg/ — Homepage-BFS vs enumeration (ARG). • fase_b2_validasi_artikel/ — Stratified sample + dual-rubric agreement summary (no full-text bodies). • fase_b3_mode_revalidation/ — Freeze-host HTTP mode comparison (C1). • fase_b4_lexicon/academic_subdomain_lexicon_v1.csv — Curated lexicon: 194 unique prefixes (v1.0.0) with SHA-256 sidecar. • fase_b5_host_yield/ — Host-level yield, PHR, Gini, RCVA_P1. • figures/ — VFR–SR funnel-consistency scatter (n1 = 21). • DATA_RELEASE_POLICY.md — Policy A (full transparency) active; Policy B documented as fallback. Replication scope. Aggregate Phase-1 statistics (n1 = 21) and full funnel reconstruction for three purposive case studies (institute / university / polytechnic). Release policy note. Campus1–Campus22 are stable analytical labels. Filenames containing “_anonymized” are legacy and do not imply identity concealment.