Matches in SemOpenAlex for { <https://semopenalex.org/work/W3143761004> ?p ?o ?g. }
- W3143761004 abstract "The right to be forgotten states that a data subject has the right to erase their data from an entity storing it. In the context of machine learning (ML), it requires the ML model provider to remove the data subject's data from the training set used to build the ML model, a process known as textit{machine unlearning}. While straightforward and legitimate, retraining the ML model from scratch upon receiving unlearning requests incurs high computational overhead when the training set is large. To address this issue, a number of approximate algorithms have been proposed in the domain of image and text data, among which SISA is the state-of-the-art solution. It randomly partitions the training set into multiple shards and trains a constituent model for each shard. However, directly applying SISA to the graph data can severely damage the graph structural information, and thereby the resulting ML model utility. In this paper, we propose GraphEraser, a novel machine unlearning method tailored to graph data. Its contributions include two novel graph partition algorithms, and a learning-based aggregation method. We conduct extensive experiments on five real-world datasets to illustrate the unlearning efficiency and model utility of GraphEraser. We observe that GraphEraser achieves 2.06$times$ (small dataset) to 35.94$times$ (large dataset) unlearning time improvement compared to retraining from scratch. On the other hand, GraphEraser achieves up to $62.5%$ higher F1 score than that of random partitioning. In addition, our proposed learning-based aggregation method achieves up to $112%$ higher F1 score than that of the majority vote aggregation." @default.
- W3143761004 created "2021-04-13" @default.
- W3143761004 creator A5004014878 @default.
- W3143761004 creator A5005741793 @default.
- W3143761004 creator A5012718640 @default.
- W3143761004 creator A5044420496 @default.
- W3143761004 creator A5053332520 @default.
- W3143761004 creator A5071833362 @default.
- W3143761004 date "2021-03-27" @default.
- W3143761004 modified "2023-09-26" @default.
- W3143761004 title "Graph Unlearning." @default.
- W3143761004 cites W1488996941 @default.
- W3143761004 cites W1692247819 @default.
- W3143761004 cites W1969198379 @default.
- W3143761004 cites W1971421925 @default.
- W3143761004 cites W2051267297 @default.
- W3143761004 cites W2132202037 @default.
- W3143761004 cites W2161160262 @default.
- W3143761004 cites W2164998314 @default.
- W3143761004 cites W2187089797 @default.
- W3143761004 cites W2535690855 @default.
- W3143761004 cites W2604366058 @default.
- W3143761004 cites W2742503211 @default.
- W3143761004 cites W2748789698 @default.
- W3143761004 cites W2753783305 @default.
- W3143761004 cites W2809583854 @default.
- W3143761004 cites W2900470550 @default.
- W3143761004 cites W2904449562 @default.
- W3143761004 cites W2962711740 @default.
- W3143761004 cites W2962756421 @default.
- W3143761004 cites W2962763344 @default.
- W3143761004 cites W2962767366 @default.
- W3143761004 cites W2963303354 @default.
- W3143761004 cites W2963312446 @default.
- W3143761004 cites W2963378725 @default.
- W3143761004 cites W2963456518 @default.
- W3143761004 cites W2963555845 @default.
- W3143761004 cites W2963744840 @default.
- W3143761004 cites W2963857521 @default.
- W3143761004 cites W2963858333 @default.
- W3143761004 cites W2963920355 @default.
- W3143761004 cites W2963984147 @default.
- W3143761004 cites W2964015378 @default.
- W3143761004 cites W2964082701 @default.
- W3143761004 cites W2964318098 @default.
- W3143761004 cites W2964321699 @default.
- W3143761004 cites W2971124187 @default.
- W3143761004 cites W2978484973 @default.
- W3143761004 cites W2995525544 @default.
- W3143761004 cites W2997156230 @default.
- W3143761004 cites W3004443750 @default.
- W3143761004 cites W3005104128 @default.
- W3143761004 cites W3007299160 @default.
- W3143761004 cites W3009148757 @default.
- W3143761004 cites W3010216907 @default.
- W3143761004 cites W3013294133 @default.
- W3143761004 cites W3035556513 @default.
- W3143761004 cites W3035616549 @default.
- W3143761004 cites W3035644192 @default.
- W3143761004 cites W3081306295 @default.
- W3143761004 cites W3084428871 @default.
- W3143761004 cites W3084563952 @default.
- W3143761004 cites W3093589247 @default.
- W3143761004 cites W3095155273 @default.
- W3143761004 cites W3095461782 @default.
- W3143761004 cites W3103973800 @default.
- W3143761004 cites W3104307750 @default.
- W3143761004 cites W3113261720 @default.
- W3143761004 cites W3154155772 @default.
- W3143761004 cites W3167334189 @default.
- W3143761004 cites W3189551349 @default.
- W3143761004 cites W9657784 @default.
- W3143761004 cites W3156148409 @default.
- W3143761004 hasPublicationYear "2021" @default.
- W3143761004 type Work @default.
- W3143761004 sameAs 3143761004 @default.
- W3143761004 citedByCount "0" @default.
- W3143761004 crossrefType "posted-content" @default.
- W3143761004 hasAuthorship W3143761004A5004014878 @default.
- W3143761004 hasAuthorship W3143761004A5005741793 @default.
- W3143761004 hasAuthorship W3143761004A5012718640 @default.
- W3143761004 hasAuthorship W3143761004A5044420496 @default.
- W3143761004 hasAuthorship W3143761004A5053332520 @default.
- W3143761004 hasAuthorship W3143761004A5071833362 @default.
- W3143761004 hasConcept C111919701 @default.
- W3143761004 hasConcept C11413529 @default.
- W3143761004 hasConcept C114614502 @default.
- W3143761004 hasConcept C119857082 @default.
- W3143761004 hasConcept C132525143 @default.
- W3143761004 hasConcept C144133560 @default.
- W3143761004 hasConcept C154945302 @default.
- W3143761004 hasConcept C155202549 @default.
- W3143761004 hasConcept C176032523 @default.
- W3143761004 hasConcept C2778712577 @default.
- W3143761004 hasConcept C2781235140 @default.
- W3143761004 hasConcept C33923547 @default.
- W3143761004 hasConcept C41008148 @default.
- W3143761004 hasConcept C42812 @default.
- W3143761004 hasConcept C51632099 @default.
- W3143761004 hasConcept C74197172 @default.