{"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/a-shared-latent-space-matrix-factorisation","title":"A shared latent space matrix factorisation method for recommending new trial evidence for systematic review updates","arxiv_id":"1709.06758","date":"2018-02-27","proceeding":null,"authors":["Surian Didi","Dunn Adam G.","Orenstein Liat","Bashir Rabia","Coiera Enrico","Bourgeois Florence T."],"abstract":"Clinical trial registries can be used to monitor the production of trial\nevidence and signal when systematic reviews become out of date. However, this\nuse has been limited to date due to the extensive manual review required to\nsearch for and screen relevant trial registrations. Our aim was to evaluate a\nnew method that could partially automate the identification of trial\nregistrations that may be relevant for systematic review updates. We identified\n179 systematic reviews of drug interventions for type 2 diabetes, which\nincluded 537 clinical trials that had registrations in ClinicalTrials.gov. We\ntested a matrix factorisation approach that uses a shared latent space to learn\nhow to rank relevant trial registrations for each systematic review, comparing\nthe performance to document similarity to rank relevant trial registrations.\nThe two approaches were tested on a holdout set of the newest trials from the\nset of type 2 diabetes systematic reviews and an unseen set of 141 clinical\ntrial registrations from 17 updated systematic reviews published in the\nCochrane Database of Systematic Reviews. The matrix factorisation approach\noutperformed the document similarity approach with a median rank of 59 and\nrecall@100 of 60.9%, compared to a median rank of 138 and recall@100 of 42.8%\nin the document similarity baseline. In the second set of systematic reviews\nand their updates, the highest performing approach used document similarity and\ngave a median rank of 67 (recall@100 of 62.9%). The proposed method was useful\nfor ranking trial registrations to reduce the manual workload associated with\nfinding relevant trials for systematic review updates. The results suggest that\nthe approach could be used as part of a semi-automated pipeline for monitoring\npotentially new evidence for inclusion in a review update.","url_abs":"http://arxiv.org/abs/1709.06758v4","url_pdf":"http://arxiv.org/pdf/1709.06758v4.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":"a-shared-latent-space-matrix-factorisation","repo_url":"https://github.com/dsurian/matfac","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"holdout-set","task_name":"Holdout Set"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}