{"title":"工具/数据集论文:使用条件表格GAN生成现实ABAC数据","authors":"Ritwik Rai, S. Sural","doi":"10.1145/3577923.3583635","DOIUrl":null,"url":null,"abstract":"Attribute-based Access Control (ABAC) is increasingly being used in a wide variety of applications that include cloud services, IoT, smart homes, healthcare and several others. Conducting systematic and reproducible experiments with benchmark realistic datasets, however, still remains a challenge. To address this shortcoming, in this paper we introduce a method called ConGRASS (Conditional Tabular GAN for Realistic ABAC Simulation Studies) for generating large ABAC datasets. Starting with a given real world dataset of (potentially) limited size, we first train a conditional tabular generative adversarial network for learning its distribution. The trained model is used to generate realistic datasets of arbitrarily large sizes having distribution similar to the original dataset. ConGRASS has been implemented as a free to use web-based tool in which a user can choose the name of a listed real dataset along with the desired dataset size. A CSV file containing ABAC data is generated as output. Extensive evaluation shows the ability of the model to faithfully learn the statistical properties of the selected real data. When such a dataset is used in an actual problem, significant improvement in performance is achieved, proving the utility of ConGRASS.","PeriodicalId":387479,"journal":{"name":"Proceedings of the Thirteenth ACM Conference on Data and Application Security and Privacy","volume":"18 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2023-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Tool/Dataset Paper: Realistic ABAC Data Generation using Conditional Tabular GAN\",\"authors\":\"Ritwik Rai, S. Sural\",\"doi\":\"10.1145/3577923.3583635\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Attribute-based Access Control (ABAC) is increasingly being used in a wide variety of applications that include cloud services, IoT, smart homes, healthcare and several others. Conducting systematic and reproducible experiments with benchmark realistic datasets, however, still remains a challenge. To address this shortcoming, in this paper we introduce a method called ConGRASS (Conditional Tabular GAN for Realistic ABAC Simulation Studies) for generating large ABAC datasets. Starting with a given real world dataset of (potentially) limited size, we first train a conditional tabular generative adversarial network for learning its distribution. The trained model is used to generate realistic datasets of arbitrarily large sizes having distribution similar to the original dataset. ConGRASS has been implemented as a free to use web-based tool in which a user can choose the name of a listed real dataset along with the desired dataset size. A CSV file containing ABAC data is generated as output. Extensive evaluation shows the ability of the model to faithfully learn the statistical properties of the selected real data. When such a dataset is used in an actual problem, significant improvement in performance is achieved, proving the utility of ConGRASS.\",\"PeriodicalId\":387479,\"journal\":{\"name\":\"Proceedings of the Thirteenth ACM Conference on Data and Application Security and Privacy\",\"volume\":\"18 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2023-04-24\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the Thirteenth ACM Conference on Data and Application Security and Privacy\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3577923.3583635\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the Thirteenth ACM Conference on Data and Application Security and Privacy","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3577923.3583635","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Tool/Dataset Paper: Realistic ABAC Data Generation using Conditional Tabular GAN
Attribute-based Access Control (ABAC) is increasingly being used in a wide variety of applications that include cloud services, IoT, smart homes, healthcare and several others. Conducting systematic and reproducible experiments with benchmark realistic datasets, however, still remains a challenge. To address this shortcoming, in this paper we introduce a method called ConGRASS (Conditional Tabular GAN for Realistic ABAC Simulation Studies) for generating large ABAC datasets. Starting with a given real world dataset of (potentially) limited size, we first train a conditional tabular generative adversarial network for learning its distribution. The trained model is used to generate realistic datasets of arbitrarily large sizes having distribution similar to the original dataset. ConGRASS has been implemented as a free to use web-based tool in which a user can choose the name of a listed real dataset along with the desired dataset size. A CSV file containing ABAC data is generated as output. Extensive evaluation shows the ability of the model to faithfully learn the statistical properties of the selected real data. When such a dataset is used in an actual problem, significant improvement in performance is achieved, proving the utility of ConGRASS.