{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QFB4EV6BE3XNU7FNAYS53XUZCI","short_pith_number":"pith:QFB4EV6B","schema_version":"1.0","canonical_sha256":"8143c257c126eeda7cad0625ddde991220745679482e8fcc13fd34a17a4efc15","source":{"kind":"arxiv","id":"2604.23025","version":2},"attestation_state":"computed","paper":{"title":"Self-Supervised Learning for Android Malware Detection on a Time-Stamped Dataset","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Self-supervised pre-training on a time-verified Android app dataset delivers 98% accuracy and 89% F1 for malware detection under realistic time constraints.","cross_cats":["cs.LG"],"primary_cat":"cs.CR","authors_text":"Annan Fu, Hao Pei, Maryam Tanha","submitted_at":"2026-04-24T21:24:48Z","abstract_excerpt":"Android malware detectors built with machine learning often suffer from temporal bias: models are trained and evaluated without respecting apps' actual release times, inflating accuracy and weakening real-world robustness. We address this by constructing a time-stamped dataset of benign and malicious Android apps and introducing a timestamp-verification procedure to ensure temporal accuracy. We then propose a detection framework that uses Bootstrap Your Own Latent (BYOL) for self-supervised pre-training to learn obfuscation-resilient representations, followed by supervised classification. Unde"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2604.23025","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2026-04-24T21:24:48Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"7edf822aefd7da0515df2cb3eb94ed9289aac5c94a7d25fc3afaccfd99845172","abstract_canon_sha256":"351388df72fef5079dfd4805e8f473ea06e7c364813b60da3042a7cb321b2760"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-08T01:04:05.562546Z","signature_b64":"bsYXfaZ8/PaZbAWyYtr6d91aTLAs6Csuy0rFwIYeCBYarbq6aDEbdBrNNrKZkPtiD6Bd7zMYs+bt2nDOqCfQAQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8143c257c126eeda7cad0625ddde991220745679482e8fcc13fd34a17a4efc15","last_reissued_at":"2026-06-08T01:04:05.561629Z","signature_status":"signed_v1","first_computed_at":"2026-06-08T01:04:05.561629Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Self-Supervised Learning for Android Malware Detection on a Time-Stamped Dataset","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Self-supervised pre-training on a time-verified Android app dataset delivers 98% accuracy and 89% F1 for malware detection under realistic time constraints.","cross_cats":["cs.LG"],"primary_cat":"cs.CR","authors_text":"Annan Fu, Hao Pei, Maryam Tanha","submitted_at":"2026-04-24T21:24:48Z","abstract_excerpt":"Android malware detectors built with machine learning often suffer from temporal bias: models are trained and evaluated without respecting apps' actual release times, inflating accuracy and weakening real-world robustness. We address this by constructing a time-stamped dataset of benign and malicious Android apps and introducing a timestamp-verification procedure to ensure temporal accuracy. We then propose a detection framework that uses Bootstrap Your Own Latent (BYOL) for self-supervised pre-training to learn obfuscation-resilient representations, followed by supervised classification. Unde"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Under time-aware evaluation, the method attains 98% accuracy and 89% F1.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The timestamp-verification procedure produces a dataset whose temporal distribution matches real-world app release patterns and that BYOL pre-training yields obfuscation-resilient representations sufficient for the downstream classification task.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"A verified time-stamped Android app dataset combined with BYOL self-supervised pre-training yields 98% accuracy and 89% F1 under time-aware malware detection evaluation.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Self-supervised pre-training on a time-verified Android app dataset delivers 98% accuracy and 89% F1 for malware detection under realistic time constraints.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"1c17873e1c06a90526767f4f380e25592167f5ce6b3c40b822b561d39e32faea"},"source":{"id":"2604.23025","kind":"arxiv","version":2},"verdict":{"id":"80b9449a-f23a-49e4-950e-300d92874527","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-08T11:20:57.625798Z","strongest_claim":"Under time-aware evaluation, the method attains 98% accuracy and 89% F1.","one_line_summary":"A verified time-stamped Android app dataset combined with BYOL self-supervised pre-training yields 98% accuracy and 89% F1 under time-aware malware detection evaluation.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The timestamp-verification procedure produces a dataset whose temporal distribution matches real-world app release patterns and that BYOL pre-training yields obfuscation-resilient representations sufficient for the downstream classification task.","pith_extraction_headline":"Self-supervised pre-training on a time-verified Android app dataset delivers 98% accuracy and 89% F1 for malware detection under realistic time constraints."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.23025/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"ai_meta_artifact","ran_at":"2026-05-21T09:40:55.301583Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:33:02.745060Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"3881be0b9ea569d97e57d143ee49bdd26e91c12593fa0001fdd8a1a444d56cd7"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2604.23025","created_at":"2026-06-08T01:04:05.561736+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.23025v2","created_at":"2026-06-08T01:04:05.561736+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.23025","created_at":"2026-06-08T01:04:05.561736+00:00"},{"alias_kind":"pith_short_12","alias_value":"QFB4EV6BE3XN","created_at":"2026-06-08T01:04:05.561736+00:00"},{"alias_kind":"pith_short_16","alias_value":"QFB4EV6BE3XNU7FN","created_at":"2026-06-08T01:04:05.561736+00:00"},{"alias_kind":"pith_short_8","alias_value":"QFB4EV6B","created_at":"2026-06-08T01:04:05.561736+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI","json":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI.json","graph_json":"https://pith.science/api/pith-number/QFB4EV6BE3XNU7FNAYS53XUZCI/graph.json","events_json":"https://pith.science/api/pith-number/QFB4EV6BE3XNU7FNAYS53XUZCI/events.json","paper":"https://pith.science/paper/QFB4EV6B"},"agent_actions":{"view_html":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI","download_json":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI.json","view_paper":"https://pith.science/paper/QFB4EV6B","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.23025&json=true","fetch_graph":"https://pith.science/api/pith-number/QFB4EV6BE3XNU7FNAYS53XUZCI/graph.json","fetch_events":"https://pith.science/api/pith-number/QFB4EV6BE3XNU7FNAYS53XUZCI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI/action/storage_attestation","attest_author":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI/action/author_attestation","sign_citation":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI/action/citation_signature","submit_replication":"https://pith.science/pith/QFB4EV6BE3XNU7FNAYS53XUZCI/action/replication_record"}},"created_at":"2026-06-08T01:04:05.561736+00:00","updated_at":"2026-06-08T01:04:05.561736+00:00"}