{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/learning-multi-touch-conversion-attribution","title":"Learning Multi-touch Conversion Attribution with Dual-attention Mechanisms for Online Advertising","arxiv_id":"1808.03737","date":"2018-08-11","proceeding":null,"authors":["Kan Ren","Yuchen Fang","Wei-Nan Zhang","Shuhao Liu","Jiajun Li","Ya zhang","Yong Yu","Jun Wang"],"abstract":"In online advertising, the Internet users may be exposed to a sequence of\ndifferent ad campaigns, i.e., display ads, search, or referrals from multiple\nchannels, before led up to any final sales conversion and transaction. For both\ncampaigners and publishers, it is fundamentally critical to estimate the\ncontribution from ad campaign touch-points during the customer journey\n(conversion funnel) and assign the right credit to the right ad exposure\naccordingly. However, the existing research on the multi-touch attribution\nproblem lacks a principled way of utilizing the users' pre-conversion actions\n(i.e., clicks), and quite often fails to model the sequential patterns among\nthe touch points from a user's behavior data. To make it worse, the current\nindustry practice is merely employing a set of arbitrary rules as the\nattribution model, e.g., the popular last-touch model assigns 100% credit to\nthe final touch-point regardless of actual attributions. In this paper, we\npropose a Dual-attention Recurrent Neural Network (DARNN) for the multi-touch\nattribution problem. It learns the attribution values through an attention\nmechanism directly from the conversion estimation objective. To achieve this,\nwe utilize sequence-to-sequence prediction for user clicks, and combine both\npost-view and post-click attribution patterns together for the final conversion\nestimation. To quantitatively benchmark attribution models, we also propose a\nnovel yet practical attribution evaluation scheme through the proxy of budget\nallocation (under the estimated attributions) over ad channels. The\nexperimental results on two real datasets demonstrate the significant\nperformance gains of our attribution model against the state of the art.","url_abs":"http://arxiv.org/abs/1808.03737v2","url_pdf":"http://arxiv.org/pdf/1808.03737v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"learning-multi-touch-conversion-attribution","repo_url":"https://github.com/rk2900/deep-conv-attr","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1808.03737","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}