{"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/learning-edge-properties-in-graphs-from-path","title":"Learning Edge Properties in Graphs from Path Aggregations","arxiv_id":"1903.04613","date":"2019-03-11","proceeding":null,"authors":["Rakshit Agrawal","Luca de Alfaro"],"abstract":"Graph edges, along with their labels, can represent information of\nfundamental importance, such as links between web pages, friendship between\nusers, the rating given by users to other users or items, and much more. We\nintroduce LEAP, a trainable, general framework for predicting the presence and\nproperties of edges on the basis of the local structure, topology, and labels\nof the graph. The LEAP framework is based on the exploration and\nmachine-learning aggregation of the paths connecting nodes in a graph. We\nprovide several methods for performing the aggregation phase by training path\naggregators, and we demonstrate the flexibility and generality of the framework\nby applying it to the prediction of links and user ratings in social networks.\n  We validate the LEAP framework on two problems: link prediction, and user\nrating prediction. On eight large datasets, among which the arXiv collaboration\nnetwork, the Yeast protein-protein interaction, and the US airlines routes\nnetwork, we show that the link prediction performance of LEAP is at least as\ngood as the current state of the art methods, such as SEAL and WLNM. Next, we\nconsider the problem of predicting user ratings on other users: this problem is\nknown as the edge-weight prediction problem in weighted signed networks (WSN).\nOn Bitcoin networks, and Wikipedia RfA, we show that LEAP performs consistently\nbetter than the Fairness & Goodness based regression models, varying the amount\nof training edges between 10 to 90%. These examples demonstrate that LEAP, in\nspite of its generality, can match or best the performance of approaches that\nhave been especially crafted to solve very specific edge prediction problems.","url_abs":"http://arxiv.org/abs/1903.04613v1","url_pdf":"http://arxiv.org/pdf/1903.04613v1.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":"learning-edge-properties-in-graphs-from-path","repo_url":"https://github.com/rakshit-agrawal/LEAP","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"fairness","task_name":"Fairness"},{"task_slug":"link-prediction","task_name":"Link Prediction"},{"task_slug":"prediction","task_name":"Prediction"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}