{"paper":{"title":"ALM-MTA:Front-Door Causal Multi-Touch Attribution Method for Creator-Ecosystem Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"Front-door identification with an adversarially learned mediator enables accurate multi-touch attribution from observational recommendation logs.","cross_cats":[],"primary_cat":"cs.SI","authors_text":"Han Li, Hu Liu, Jian Liang, Kun Gai, Luyao Xia, Yuguang Liu, Zhangxi Yan","submitted_at":"2026-05-09T11:04:18Z","abstract_excerpt":"Consumption Drives Production (CDP) on social platforms aims to deliver interpretable incentive signals for creator ecosystem building and resource utilization improvement, which strongly relies on attribution. In large-scale and complex recommendation systems, the absence of accurate labels together with unobserved confounding renders backdoor adjustments alone insufficient for reliable attribution. To address these problems, we propose Adversarial Learning Mediator based Multi-Touch Attribution (ALM-MTA), an extensible causal framework that leverages front-door identification with an adversa"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"ALM-MTA increases DAU by 0.04% and daily active creators by 0.6%, with unit exposure efficiency increased by 670%. On causal utility, ALM-MTA achieves higher grouped AUUC than the SOTA in every propensity bucket, with a maximum gain of 0.070. In terms of accuracy, ALM-MTA improves upload AUC by 40% compared to SOTA.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The adversarially learned mediator successfully distills outcome information to strengthen the causal pathway from treatment to outcome while eliminating shortcut leakage, and that contrastive learning on high-match pairs ensures positivity without introducing selection bias in the large treatment space.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"ALM-MTA uses front-door causal inference with an adversarially trained mediator and contrastive learning to improve multi-touch attribution, reporting gains in DAU, creator activity, exposure efficiency, AUUC, and upload AUC on a 400M DAU platform.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Front-door identification with an adversarially learned mediator enables accurate multi-touch attribution from observational recommendation logs.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"efcdb1fd6c5f2af49e1a723a3af103eecfc73c85210eb5ed770bc4418adc8b8c"},"source":{"id":"2605.08881","kind":"arxiv","version":2},"verdict":{"id":"e69e107f-a903-4b95-b01e-d72d570092cb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-12T01:05:39.748434Z","strongest_claim":"ALM-MTA increases DAU by 0.04% and daily active creators by 0.6%, with unit exposure efficiency increased by 670%. On causal utility, ALM-MTA achieves higher grouped AUUC than the SOTA in every propensity bucket, with a maximum gain of 0.070. In terms of accuracy, ALM-MTA improves upload AUC by 40% compared to SOTA.","one_line_summary":"ALM-MTA uses front-door causal inference with an adversarially trained mediator and contrastive learning to improve multi-touch attribution, reporting gains in DAU, creator activity, exposure efficiency, AUUC, and upload AUC on a 400M DAU platform.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The adversarially learned mediator successfully distills outcome information to strengthen the causal pathway from treatment to outcome while eliminating shortcut leakage, and that contrastive learning on high-match pairs ensures positivity without introducing selection bias in the large treatment space.","pith_extraction_headline":"Front-door identification with an adversarially learned mediator enables accurate multi-touch attribution from observational recommendation logs."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.08881/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"claim_evidence","ran_at":"2026-05-20T08:42:01.978321Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T20:42:27.427743Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_title_agreement","ran_at":"2026-05-19T14:01:21.572223Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T10:42:06.978515Z","status":"completed","version":"1.0.0","findings_count":0}],"snapshot_sha256":"609821cb559d09d79eb1cce2977fc331309d6cba7038a57ae72fa0f1e225ace5"},"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"}