[{"publication_identifier":{"issn":["0004-3702"]},"language":[{"iso":"eng"}],"date_published":"2024-09-01T00:00:00Z","isi":1,"abstract":[{"text":"In a delegation problem, a principal P with commitment power tries to pick one out of 𝑛 options.\r\nEach option is drawn independently from a known distribution. Instead of inspecting the options\r\nherself, P delegates the information acquisition to a rational and self-interested agent A. After\r\ninspection, A proposes one of the options, and P can accept or reject.\r\nDelegation is a classic setting in economic information design with many prominent applications,\r\nbut the computational problems are only poorly understood. In this paper, we study a natural\r\nonline variant of delegation, in which the agent searches through the options in an online fashion.\r\nFor each option, he has to irrevocably decide if he wants to propose the current option or discard\r\nit, before seeing information on the next option(s). How can we design algorithms for P that\r\napproximate the utility of her best option in hindsight?\r\nWe show that in general P can obtain a Θ(1∕𝑛)-approximation and extend this result to ratios\r\nof Θ(𝑘∕𝑛) in case (1) A has a lookahead of 𝑘 rounds, or (2) A can propose up to 𝑘 different\r\noptions. We provide fine-grained bounds independent of 𝑛 based on three parameters. If the ratio\r\nof maximum and minimum utility for A is bounded by a factor 𝛼, we obtain an Ω(loglog 𝛼∕ log 𝛼)-\r\napproximation algorithm, and we show that this is best possible. Additionally, if P cannot\r\ndistinguish options with the same value for herself, we show that ratios polynomial in 1∕𝛼 cannot\r\nbe avoided. If there are at most 𝛽 different utility values for A, we show a Θ(1∕𝛽)-approximation.\r\nIf the utilities of P and A for each option are related by a factor 𝛾, we obtain an Ω(1∕ log 𝛾)-\r\napproximation, where 𝑂(log log 𝛾∕ log 𝛾) is best possible.","lang":"eng"}],"year":"2024","citation":{"short":"P. Braun, N. Hahn, M. Hoefer, C. Schecker, Artificial Intelligence 334 (2024).","ieee":"P. Braun, N. Hahn, M. Hoefer, and C. Schecker, “Delegated online search,” <i>Artificial Intelligence</i>, vol. 334. Elsevier, 2024.","ama":"Braun P, Hahn N, Hoefer M, Schecker C. Delegated online search. <i>Artificial Intelligence</i>. 2024;334. doi:<a href=\"https://doi.org/10.1016/j.artint.2024.104171\">10.1016/j.artint.2024.104171</a>","ista":"Braun P, Hahn N, Hoefer M, Schecker C. 2024. Delegated online search. Artificial Intelligence. 334, 104171.","chicago":"Braun, Pirmin, Niklas Hahn, Martin Hoefer, and Conrad Schecker. “Delegated Online Search.” <i>Artificial Intelligence</i>. Elsevier, 2024. <a href=\"https://doi.org/10.1016/j.artint.2024.104171\">https://doi.org/10.1016/j.artint.2024.104171</a>.","apa":"Braun, P., Hahn, N., Hoefer, M., &#38; Schecker, C. (2024). Delegated online search. <i>Artificial Intelligence</i>. Elsevier. <a href=\"https://doi.org/10.1016/j.artint.2024.104171\">https://doi.org/10.1016/j.artint.2024.104171</a>","mla":"Braun, Pirmin, et al. “Delegated Online Search.” <i>Artificial Intelligence</i>, vol. 334, 104171, Elsevier, 2024, doi:<a href=\"https://doi.org/10.1016/j.artint.2024.104171\">10.1016/j.artint.2024.104171</a>."},"intvolume":"       334","article_processing_charge":"Yes (in subscription journal)","publisher":"Elsevier","file_date_updated":"2025-01-09T10:45:24Z","file":[{"checksum":"f02a56bc7ea88f41fcc68968e4ceddf3","file_size":772226,"date_updated":"2025-01-09T10:45:24Z","file_id":"18806","creator":"dernst","relation":"main_file","file_name":"2024_ArtificialIntelligence_Braun.pdf","content_type":"application/pdf","success":1,"date_created":"2025-01-09T10:45:24Z","access_level":"open_access"}],"OA_type":"hybrid","title":"Delegated online search","day":"01","tmp":{"name":"Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)","image":"/images/cc_by.png","legal_code_url":"https://creativecommons.org/licenses/by/4.0/legalcode","short":"CC BY (4.0)"},"external_id":{"isi":["001260448100001"],"arxiv":["2203.01084"]},"article_number":"104171","scopus_import":"1","OA_place":"publisher","department":[{"_id":"MoHe"}],"arxiv":1,"doi":"10.1016/j.artint.2024.104171","quality_controlled":"1","author":[{"last_name":"Braun","first_name":"Pirmin","full_name":"Braun, Pirmin"},{"last_name":"Hahn","id":"0a01c7b2-b823-11ed-9928-cc3f874f9ffd","first_name":"Niklas","full_name":"Hahn, Niklas"},{"first_name":"Martin","full_name":"Hoefer, Martin","last_name":"Hoefer"},{"last_name":"Schecker","first_name":"Conrad","full_name":"Schecker, Conrad"}],"ddc":["000"],"article_type":"original","corr_author":"1","_id":"17188","date_created":"2024-06-30T22:01:05Z","publication_status":"published","acknowledgement":"Hahn gratefully acknowledges the support of GIF grant I-1419-118.4/2017. Hoefer gratefully acknowledges the support of GIF grant I-1419-118.4/2017, DFG Research Unit ADYN (project number 411362735), and DFG grant Ho 3831/9-1 (project number 514505843).","type":"journal_article","has_accepted_license":"1","oa":1,"volume":334,"date_updated":"2025-09-08T08:00:42Z","status":"public","oa_version":"Published Version","user_id":"317138e5-6ab7-11ef-aa6d-ffef3953e345","publication":"Artificial Intelligence","month":"09"}]