{"id":889,"date":"2025-09-09T10:29:41","date_gmt":"2025-09-09T10:29:41","guid":{"rendered":"https:\/\/biconnect.psicologia.unimib.it\/?p=889"},"modified":"2025-09-09T18:36:55","modified_gmt":"2025-09-09T18:36:55","slug":"podcast-release-insights-from-our-work-on-data-skeptic","status":"publish","type":"post","link":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/podcast-release-insights-from-our-work-on-data-skeptic\/","title":{"rendered":"Podcast Release: Insights from Our Paper &#8220;Why am I seeing this?&#8221; on Data Skeptic"},"content":{"rendered":"\n<p>We recently joined the <a href=\"https:\/\/dataskeptic.com\/podcast\/episodes\/why-am-i-seeing-this\" target=\"_blank\" rel=\"noopener\">Data Skeptic<\/a> episode <em>\u201cWhy Am I Seeing This?\u201d<\/em> to discuss our research on identifying and understanding hidden social-media recommender systems. The conversation walks through our approach, why exposure data is often missing, and what that means for researchers, platforms, and policy.<\/p>\n\n\n\n<figure class=\"wp-block-embed is-type-rich is-provider-spotify wp-block-embed-spotify\"><div class=\"wp-block-embed__wrapper\">\n<iframe title=\"Spotify Embed: Why Am I Seeing This?\" style=\"border-radius: 12px\" width=\"100%\" height=\"152\" frameborder=\"0\" allowfullscreen allow=\"autoplay; clipboard-write; encrypted-media; fullscreen; picture-in-picture\" loading=\"lazy\" src=\"https:\/\/open.spotify.com\/embed\/episode\/3aqVrnUnHjy2VpYTdeM3rC?si=7adefdd611544385&#038;nd=1&#038;dlsi=1b08b991692b44b5&#038;utm_source=oembed\"><\/iframe>\n<\/div><\/figure>\n\n\n\n<div style=\"height:50px\" aria-hidden=\"true\" class=\"wp-block-spacer\"><\/div>\n\n\n\n<h2 class=\"wp-block-heading\">Why this matters<\/h2>\n\n\n\n<p>Recommenders quietly shape what we see\u2014and what we believe. Because platforms rarely expose the underlying algorithms or full exposure logs, it\u2019s hard to audit their real effects on behavior, polarization, or misinformation.<br>Our work tackles this gap by inferring the <em>kind<\/em> of recommender likely at play from observable network signals and user actions, without needing access to users\u2019 impression data. That lets us reason about downstream impacts and compare \u201cwhat if\u201d scenarios responsibly.<\/p>\n\n\n\n<h2 class=\"wp-block-heading\">What we contributed (and talked about on the show)<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Automatic Recommender Recognition<\/strong> with Graph Neural Networks (GNNs): we train a <em>Recommender-Neutral User<\/em> model to strip out platform-specific effects, then generate <em>recommender-hypothesis<\/em> synthetic datasets to see which algorithm best explains the observed world.<\/li>\n\n\n\n<li><strong>Academic networks as an ethical testbed:<\/strong> we use scholarly collaboration networks as a proxy to study recommendation dynamics and run controlled \u201cinfosphere vs. behavior\u201d simulations with GNNs.<\/li>\n\n\n\n<li><strong>From recognition to mitigation:<\/strong> in parallel, our collaborators design <em>collective-well-being-aware<\/em> recommenders and a <strong>Virtual Learning Companion<\/strong> that coaches teens to handle toxic and misleading content\u2014pairing automated detectors with educator input to improve community well-being.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">A bit of motivation from the referenced work<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Opacity shouldn\u2019t stall safety:<\/strong> if we can <em>recognize<\/em> recommender behavior from traces in the network, we can study systemic effects even when impressions are missing\u2014turning a long-standing data-access problem into a modeling problem.<\/li>\n\n\n\n<li><strong>Simulate before you legislate:<\/strong> synthetic, hypothesis-driven datasets let researchers and policymakers ask counterfactuals (\u201cWhat if the platform used X instead of Y?\u201d) without experimenting on real users.<\/li>\n\n\n\n<li><strong>Well-being as an objective:<\/strong> when we optimize only for engagement, platforms can amplify popularity and outrage biases; framing recommendation around <em>collective well-being<\/em> offers a concrete alternative.<\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">Papers &amp; resources<\/h2>\n\n\n\n<ul class=\"wp-block-list\">\n<li><a href=\"https:\/\/arxiv.org\/abs\/2504.11000\" target=\"_blank\" rel=\"noopener\"><strong>Why am I seeing this? Towards recognizing social media recommender systems with missing recommendations<\/strong><\/a> (2025).<\/li>\n\n\n\n<li><a href=\"https:\/\/arxiv.org\/abs\/2410.04552\" target=\"_blank\" rel=\"noopener\"><strong>Modeling Social Media Recommendation Impacts Using Academic Networks: A Graph Neural Network Approach<\/strong><\/a> (2024).<\/li>\n\n\n\n<li><a href=\"https:\/\/www.frontiersin.org\/journals\/artificial-intelligence\/articles\/10.3389\/frai.2022.654930\/full\" target=\"_blank\" rel=\"noopener\"><strong>Challenging Social Media Threats using Collective Well-being-Aware Recommendation Algorithms and an Educational Virtual Companion<\/strong><\/a> (2023).<\/li>\n\n\n\n<li><a href=\"https:\/\/dataskeptic.com\/blog\/episodes\/2024\/why-am-i-seeing-this\" target=\"_blank\" rel=\"noopener\"><strong>Data Skeptic episode: \u201cWhy Am I Seeing This?\u201d<\/strong><\/a><\/li>\n\n\n\n<li><a href=\"https:\/\/biconnect.psicologia.unimib.it\/index.php\/courage\/\" data-type=\"page\" data-id=\"79\"><strong>COURAGE project<\/strong><\/a><\/li>\n<\/ul>\n\n\n\n<h2 class=\"wp-block-heading\">People involved<\/h2>\n\n\n\n<p>This work and the podcast are the result of a collaborative effort. The episode features <strong>Dr. Sabrina Guidotti<\/strong>, <strong>Dr. Gregor Donabauer<\/strong>, and <strong>Prof. Dimitri Ognibene<\/strong>. We also build on contributions from <strong>Prof. Udo Kruschwitz<\/strong>, <strong>Dr. Davide Taibi<\/strong>, <strong>Prof. Giuseppe Vizzari<\/strong>, <strong>Dr. Simone Somazzi<\/strong>, and <strong>Dr. Sabrina Patania<\/strong>\u2014with connected efforts through the <a href=\"https:\/\/biconnect.psicologia.unimib.it\/index.php\/courage\/\" data-type=\"page\" data-id=\"79\"><strong>COURAGE<\/strong> <\/a>initiative involving <strong>Prof. Davinia Hern\u00e1ndez Leo<\/strong>, <strong>Emily Theophilou<\/strong>, and <strong>Prof. Sabrina Eimler<\/strong>, among others.<\/p>\n\n\n\n<h3 class=\"wp-block-heading\">Contact<\/h3>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Dr. Sabrina Guidotti<\/strong> \u2014 s.guidotti2@campus.unimib.it<\/li>\n\n\n\n<li><strong>Dr. Gregor Donabauer<\/strong> \u2014 gregor.donabauer@ur.de<\/li>\n\n\n\n<li><strong>Prof. Dimitri Ognibene<\/strong> \u2014 dimitri.ognibene@unimib.it<\/li>\n\n\n\n<li><strong>Dr. Davide Taibi<\/strong> \u2014 davide.taibi@cnr.it<\/li>\n\n\n\n<li><strong>Dr. Sabrina Patania<\/strong> \u2014 sabrina.patania@unimib.it<\/li>\n\n\n\n<li><strong>Prof. Udo Kruschwitz<\/strong> \u2014 Udo.Kruschwitz@ur.de<\/li>\n\n\n\n<li><strong>Prof. Davinia Hern\u00e1ndez Leo<\/strong> \u2014 davinia.hernandez-leo@upf.edu<\/li>\n\n\n\n<li><strong>Prof. Giuseppe Vizzari<\/strong> \u2014 giuseppe.vizzari@unimib.it<\/li>\n\n\n\n<li><strong>Prof. Sabrina Eimler<\/strong> \u2014 sabrinaeimler@gmail.com<\/li>\n\n\n\n<li><strong>Dr. Emily Theophilou<\/strong> \u2014 emily.theophilou@upf.edu<\/li>\n<\/ul>\n\n\n\n<p><\/p>\n","protected":false},"excerpt":{"rendered":"<p>We recently joined the Data Skeptic episode \u201cWhy Am I Seeing This?\u201d to discuss our research on identifying and understanding hidden social-media recommender systems. The conversation walks through our approach, why exposure data is often missing, and what that means for researchers, platforms, and policy. Why this matters Recommenders quietly shape what we see\u2014and what &#8230; <a title=\"Podcast Release: Insights from Our Paper &#8220;Why am I seeing this?&#8221; on Data Skeptic\" class=\"read-more\" href=\"https:\/\/biconnect.psicologia.unimib.it\/index.php\/podcast-release-insights-from-our-work-on-data-skeptic\/\" aria-label=\"Read more about Podcast Release: Insights from Our Paper &#8220;Why am I seeing this?&#8221; on Data Skeptic\">Read more<\/a><\/p>\n","protected":false},"author":3,"featured_media":894,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"footnotes":""},"categories":[43],"tags":[],"class_list":["post-889","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-lab-news"],"acf":[],"_links":{"self":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/posts\/889","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/users\/3"}],"replies":[{"embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/comments?post=889"}],"version-history":[{"count":24,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/posts\/889\/revisions"}],"predecessor-version":[{"id":948,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/posts\/889\/revisions\/948"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/media\/894"}],"wp:attachment":[{"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/media?parent=889"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/categories?post=889"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/biconnect.psicologia.unimib.it\/index.php\/wp-json\/wp\/v2\/tags?post=889"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}