[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-160129-en":3,"doc-seo-160129-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},160129,962075114765,"Quinn","https://ap-avatar.wpscdn.com/davatar_a8503ba1806abce46bf441b54a3ca4cd",8,"Research & Report","Professional Network Matters - Connections Empower Person-Job Fit - Two-Stage WHIN and CSAGNN Approach","Online recruitment platforms rely on Person-Job Fit (PJF) models to match job seekers with suitable roles using profile and job-description signals. Existing approaches often neglect a decisive resource: users’ professional relationships within workplace networks. This paper proposes integrating professional networks into PJF via a two-stage framework: a Workplace Heterogeneous Information Network (WHIN) built with heterogeneous graph neural networks to represent entities and connections, and a Contextual Social Attention Graph Neural Network (CSAGNN) that augments missing user information with contextual signals. A job-specific attention mechanism addresses noisy networks, achieving stronger results on three real-world LinkedIn datasets than baseline methods.","Professional Network Matters: Connections Empower Person-Job Fit  \nHao Chen∗ [haochen@mail.bnu.edu.cn](haochen@mail.bnu.edu.cn)[ ](haochen@mail.bnu.edu.cn)Beijing Normal University Beijing, China  \nLun Du† [lun.du@microsoft.com](lun.du@microsoft.com)[ ](lun.du@microsoft.com)Microsoft Research Beijing, China  \nYuxuan Lu∗ [lu.yuxuan@northeastern.edu](lu.yuxuan@northeastern.edu)[ ](lu.yuxuan@northeastern.edu)Northeastern University Boston, USA  \narXiv :2401 .000 10v 1 [ cs . SI] 19 Dec 2023  \nQiang Fu  \n[qifu@microsoft.com](qifu@microsoft.com)[ ](qifu@microsoft.com)Microsoft Research Beijing, China  \nYanbin Kang  \n[ybkang@linkedin.com](ybkang@linkedin.com)[ ](ybkang@linkedin.com)LinkedIn Corporation Beijing, China  \nXu Chen  \n[xu.chen@microsoft.com](xu.chen@microsoft.com)[ ](xu.chen@microsoft.com)Microsoft Research Beijing, China  \nGuangming Lu  \n[glu@linkedin.com](glu@linkedin.com)[ ](glu@linkedin.com)LinkedIn Corporation Beijing, China  \nShi Han  \n[shihan@microsoft.com](shihan@microsoft.com)[ ](shihan@microsoft.com)Microsoft Research Beijing, China  \nZi Li  \n[zili@linkedin.com](zili@linkedin.com)[ ](zili@linkedin.com)LinkedIn Corporation Beijing, China  \nABSTRACT  \nOnline recruitment platforms typically employ Person-Job Fit models in the core service that automatically match suitable job seekers with appropriate job positions. While existing works leverage historical or contextual information, they often disregard a crucial aspect: job seekers’ social relationships in professional networks. This paper emphasizes the importance of incorporating professional networks into the Person-Job Fit model. Our innovative approach consists of two stages: (1) defining a Workplace Heterogeneous Information Network (WHIN) to capture heterogeneous knowledge, including professional connections and pre-training representations of various entities using a heterogeneous graph neural network; (2) designing a Contextual Social Attention Graph Neural Network (CSAGNN) that supplements users’ missing information with professional connections’ contextual information. We introduce a job-specific attention mechanism in CSAGNN to handle noisy professional networks, leveraging pre-trained entity representations from WHIN. We demonstrate the effectiveness of our approach through experimental evaluations conducted across three real-world recruitment datasets from LinkedIn, showing superior performance compared to baseline models.  \nKEYWORDS  \nPerson-Job Fit, Heterogeneous Information Network, Graph Neural Network  \n1 INTRODUCTION  \nWith the rapid development of the Internet, online recruitment platforms (e.g., LinkedIn1 , Indeed2 , and ZipRecruiter3) are becoming essential for recruiting and job seeking. A considerable number  \n∗ This work was done when they were interns at LinkedIn.†Corresponding author  \n1[https://www.linkedin.com](https://www.linkedin.com)  \n2[https://www.indeed.com](https://www.indeed.com)  \n3[https://www.ziprecruiter.com](https://www.ziprecruiter.com)  \nof talent profiles and job descriptions are posted on these platforms. Taking LinkedIn as an example, more than 900 million members have registered, and 90 jobs were posted every second by the first quarter of 20234 . Considering such a large number of options, Person-Job Fit (PJF) [26] has become a critical research topic for improving the efficiency of recruitment and job seeking. PersonJob Fit aims to automatically link the right talents to the right job positions according to talent competencies and job requirements.  \nPrevious works on Person-Job Fit mainly focus on leveraging two types of information, namely, (1) historical job application information and (2) textual information in profiles and job descriptions. Collaborative filtering-based methods [3, 30] are applied to capture co-apply relations between job seekers and co-applied relations between job positions in the historical application information. Manually-engineered textual features and deep language models have been wide","cbCaileYGj6NoiKa","https://ap.wps.com/l/cbCaileYGj6NoiKa","pdf",1134593,1,9,"English","en",105,"# Abstract\n# 1 Introduction\n## Person-Job Fit in online recruitment\n## Limitations of existing methods\n## Role of professional networks","[{\"question\":\"Why should professional networks be incorporated into Person-Job Fit models?\",\"answer\":\"Professional networks can bridge the gap between job seekers and relevant opportunities and can also help infer missing profile details through connections’ information.\"},{\"question\":\"What is the two-stage framework proposed in the paper?\",\"answer\":\"Stage one builds a Workplace Heterogeneous Information Network (WHIN) to capture heterogeneous knowledge and entity representations; stage two uses a Contextual Social Attention Graph Neural Network (CSAGNN) to inject contextual information from professional connections.\"},{\"question\":\"How does the method handle noise in professional networks?\",\"answer\":\"CSAGNN uses a job-specific attention mechanism so that only information useful for the target job is emphasized, while irrelevant or unhelpful connections are downweighted.\"}]","Professional Network Matters - Connections Empower Person-Job Fit - Two-Stage WHIN and CSAGNN Approach | PDF",1788050716,23,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"professional-network-matters-connections-empower-person-job-fit-two-stage-whin-and-csagnn-approach","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/professional-network-matters-connections-empower-person-job-fit-two-stage-whin-and-csagnn-approach/160129/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-30",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why should professional networks be incorporated into Person-Job Fit models?","Question",{"text":75,"@type":76},"Professional networks can bridge the gap between job seekers and relevant opportunities and can also help infer missing profile details through connections’ information.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What is the two-stage framework proposed in the paper?",{"text":80,"@type":76},"Stage one builds a Workplace Heterogeneous Information Network (WHIN) to capture heterogeneous knowledge and entity representations; stage two uses a Contextual Social Attention Graph Neural Network (CSAGNN) to inject contextual information from professional connections.",{"name":82,"@type":73,"acceptedAnswer":83},"How does the method handle noise in professional networks?",{"text":84,"@type":76},"CSAGNN uses a job-specific attention mechanism so that only information useful for the target job is emphasized, while irrelevant or unhelpful connections are downweighted.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,127,130,134],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":21,"doc_module":4,"doc_module_name":46,"category_name":124,"show_sort_weight":125,"slug":126},"Religion & Spirituality",20,"religion-spirituality",{"id":125,"doc_module":4,"doc_module_name":46,"category_name":128,"show_sort_weight":125,"slug":129},"World Cup","world-cup",{"id":131,"doc_module":4,"doc_module_name":46,"category_name":132,"show_sort_weight":131,"slug":133},10,"Lifestyle","lifestyle",{"id":135,"doc_module":4,"doc_module_name":46,"category_name":136,"show_sort_weight":106,"slug":137},19,"General","general"]