{"id":3899,"date":"2025-05-05T19:42:31","date_gmt":"2025-05-05T19:42:31","guid":{"rendered":"https:\/\/iscpif.fr\/chavalarias\/?p=3899"},"modified":"2025-07-18T21:39:28","modified_gmt":"2025-07-18T21:39:28","slug":"two-antagonistic-objectives-for-one-multi-scale-graph-clustering-framework","status":"publish","type":"post","link":"https:\/\/iscpif.fr\/chavalarias\/?p=3899","title":{"rendered":"Sci. Rep. : Two antagonistic objectives for one multi-scale graph clustering framework"},"content":{"rendered":"<div  class='flex_column av-3unid7-2ea7efb6a6fc0ac0798262cdd58295b1 av_one_half  avia-builder-el-0  el_before_av_one_half  avia-builder-el-first  first flex_column_div  '     ><p>\n<style type=\"text\/css\" data-created_by=\"avia_inline_auto\" id=\"style-css-av-mabhf32o-9616f7aef16e4276bb1973fd92306264\">\n#top .av-special-heading.av-mabhf32o-9616f7aef16e4276bb1973fd92306264{\npadding-bottom:10px;\n}\nbody .av-special-heading.av-mabhf32o-9616f7aef16e4276bb1973fd92306264 .av-special-heading-tag .heading-char{\nfont-size:25px;\n}\n.av-special-heading.av-mabhf32o-9616f7aef16e4276bb1973fd92306264 .av-subheading{\nfont-size:15px;\n}\n<\/style>\n<div  class='av-special-heading av-mabhf32o-9616f7aef16e4276bb1973fd92306264 av-special-heading-h3  avia-builder-el-1  el_before_av_textblock  avia-builder-el-first '><h3 class='av-special-heading-tag '  itemprop=\"headline\"  >Two antagonistic objectives for one multi-scale graph clustering framework<\/h3><div class=\"special-heading-border\"><div class=\"special-heading-inner-border\"><\/div><\/div><\/div><br \/>\n<section  class='av_textblock_section av-mabhfmb2-74bb4cf510650608b158a4ff5826d172 '   itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/BlogPosting\" itemprop=\"blogPost\" ><div class='avia_textblock'  itemprop=\"text\" ><p><strong>Source :<\/strong> Gaume, Bruno, Ixandra Achitouv, et David Chavalarias. 2025. \u00ab\u00a0Two Antagonistic Objectives for One Multi-Scale Graph Clustering Framework\u00a0\u00bb. <i>Scientific Reports<\/i> 15 (1): 13368. <a href=\"https:\/\/doi.org\/10.1038\/s41598-025-90454-w\">https:\/\/doi.org\/10.1038\/s41598-025-90454-w<\/a>.<\/p>\n<blockquote>\n<p>In the current state of knowledge, there is no consensus on an objective criterion for evaluating network communities as cohesive sets of nodes with the following two properties: PDC : Each community is Densely Connected; PWC : Communities are Weakly Connected to each other. This makes it difficult to conduct comparative studies between dozens of graph clustering methods proposed over more than 20 years. To fill this gap: We propose a graph clustering framework by faithfully formalizing PDC with precision and PW C with recall, which are two meaningful metrics, simple, well known and already widely used for many tasks in most sciences. The meaning of these metrics in the context of graph clustering is therefore easily interpretable by most users of real-world graphs. We show that for most graphs, these two metrics are antagonistic, i.e. there is no solution that simultaneously maximizes precision and recall. In other words, to select a clustering among the Pareto optimal solutions (clusterings such that no other clustering exist that both increases the precision and the recall) we must first make a subjective compromise, according to our needs between the two properties PDC and PW C . We then show how to use this framework to compare, even without \u2018ground truth\u2019, the performances of five hitherto incommensurable state-of-the-art clustering methods, as well as that of a new family of clustering methods inspired by our approach.<\/p>\n<\/blockquote>\n<\/div><\/section><\/p><\/div><div  class='flex_column av-bu6ij-8b4d3a54521b4fab845bf89aec236dd2 av_one_half  avia-builder-el-3  el_after_av_one_half  avia-builder-el-last  flex_column_div  '     ><p><div  class='avia-icon-list-container av-mabhiqe8-44ed79b9d899f8a3faf41bfcba532c26  avia-builder-el-4  el_before_av_image  avia-builder-el-first '><ul class='avia-icon-list avia_animate_when_almost_visible avia-icon-list-left av-iconlist-big av-mabhiqe8-44ed79b9d899f8a3faf41bfcba532c26 avia-iconlist-animate'>\n<li><a href='https:\/\/www-nature-com.inshs.bib.cnrs.fr\/articles\/s41598-025-90454-w' title='View on Scientific Reports'  target=\"_blank\"  rel=\"noopener noreferrer\" class='iconlist_icon av-mabhh4re-bfe1929f57f68cf7d75fa497f695e497 avia-font-svg_entypo-fontello avia-svg-icon avia-font-svg_entypo-fontello'><span class='av-icon-char' data-av_svg_icon='newspaper' data-av_iconset='svg_entypo-fontello'><svg version=\"1.1\" xmlns=\"http:\/\/www.w3.org\/2000\/svg\" width=\"26\" height=\"32\" viewBox=\"0 0 26 32\" preserveAspectRatio=\"xMidYMid meet\" aria-labelledby='av-svg-title-1' aria-describedby='av-svg-desc-1' role=\"graphics-symbol\" aria-hidden=\"true\">\n<title id='av-svg-title-1'>Newspaper<\/title>\n<desc id='av-svg-desc-1'>Newspaper<\/desc>\n<path d=\"M22.4 1.6q1.344 0 2.272 0.928t0.928 2.272v22.4q0 1.28-0.928 2.24t-2.272 0.96h-19.2q-1.28 0-2.24-0.96t-0.96-2.24v-22.4q0-1.344 0.96-2.272t2.24-0.928h19.2zM22.4 27.2v-22.4h-19.2v22.4h19.2zM14.4 19.2v1.6h-8v-1.6h8zM19.2 12.8v1.6h-6.4v-1.6h6.4zM12.8 11.2v-3.2h6.4v3.2h-6.4zM11.2 8v6.4h-4.8v-6.4h4.8zM9.6 16v1.6h-3.2v-1.6h3.2zM11.2 17.6v-1.6h8v1.6h-8zM19.2 22.4v1.6h-12.8v-1.6h12.8zM16 20.8v-1.6h3.2v1.6h-3.2z\"><\/path>\n<\/svg><\/span><\/a><article class=\"article-icon-entry \"  itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/BlogPosting\" itemprop=\"blogPost\" ><div class=\"iconlist_content_wrap\"><header class=\"entry-content-header\" aria-label=\"Icon: &lt;a href=&#039;https:\/\/www-nature-com.inshs.bib.cnrs.fr\/articles\/s41598-025-90454-w&#039; title=&#039;View on Scientific Reports&#039; target=&quot;_blank&quot;  rel=&quot;noopener noreferrer&quot;&gt;View on Scientific Reports&lt;\/a&gt;\"><h4 class='av_iconlist_title iconlist_title  '  itemprop=\"headline\" ><a href='https:\/\/www-nature-com.inshs.bib.cnrs.fr\/articles\/s41598-025-90454-w' title='View on Scientific Reports' target=\"_blank\"  rel=\"noopener noreferrer\">View on Scientific Reports<\/a><\/h4><\/header><div class='iconlist_content '  itemprop=\"text\" ><p>(Open Access)<\/p>\n<\/div><\/div><footer class=\"entry-footer\"><\/footer><\/article><div class=\"iconlist-timeline\"><\/div><\/li>\n<\/ul><\/div><br \/>\n\n<style type=\"text\/css\" data-created_by=\"avia_inline_auto\" id=\"style-css-av-mabhky19-583e582d22d72bf68c94e458cd4e069d\">\n.avia-image-container.av-mabhky19-583e582d22d72bf68c94e458cd4e069d img.avia_image{\nbox-shadow:none;\n}\n.avia-image-container.av-mabhky19-583e582d22d72bf68c94e458cd4e069d .av-image-caption-overlay-center{\ncolor:#ffffff;\n}\n<\/style>\n<div  class='avia-image-container av-mabhky19-583e582d22d72bf68c94e458cd4e069d av-styling- avia-align-center  avia-builder-el-5  el_after_av_iconlist  avia-builder-el-last '   itemprop=\"image\" itemscope=\"itemscope\" itemtype=\"https:\/\/schema.org\/ImageObject\" ><div class=\"avia-image-container-inner\"><div class=\"avia-image-overlay-wrap\"><img decoding=\"async\" fetchpriority=\"high\" class='wp-image-3900 avia-img-lazy-loading-not-3900 avia_image ' src=\"https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale.png\" alt='Informations of interest to define intrinsic scales as well as relevant \u2018ground-truths\u2019 for multi-scale graphs.' title='SciRep Multi-scale Fig 6'  height=\"1062\" width=\"789\"  itemprop=\"thumbnailUrl\" srcset=\"https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale.png 789w, https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale-223x300.png 223w, https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale-765x1030.png 765w, https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale-768x1034.png 768w, https:\/\/iscpif.fr\/chavalarias\/wp-content\/uploads\/sites\/4\/2025\/05\/SciRep_MultiScale-524x705.png 524w\" sizes=\"(max-width: 789px) 100vw, 789px\" \/><\/div><\/div><\/div><\/p><\/div><\/p>\n","protected":false},"excerpt":{"rendered":"","protected":false},"author":9,"featured_media":3900,"comment_status":"closed","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[32,17,3],"tags":[],"class_list":["post-3899","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","category-publication","category-toread"],"_links":{"self":[{"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/posts\/3899","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/users\/9"}],"replies":[{"embeddable":true,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=3899"}],"version-history":[{"count":4,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/posts\/3899\/revisions"}],"predecessor-version":[{"id":4009,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/posts\/3899\/revisions\/4009"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=\/wp\/v2\/media\/3900"}],"wp:attachment":[{"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=3899"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=3899"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/iscpif.fr\/chavalarias\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=3899"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}