促進因果探勘品質的區塊鏈資料分享平台;A Blockchain Data Sharing Platform for Promoting the Quality of Causal Rule Mining

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/89736

Please use this identifier to cite or link to this item: https://ir.lib.ncu.edu.tw/handle/987654321/89736

Title:	促進因果探勘品質的區塊鏈資料分享平台;A Blockchain Data Sharing Platform for Promoting the Quality of Causal Rule Mining
Authors:	楊昀傑;Yang, Yun-Chieh
Contributors:	資訊工程學系
Keywords:	智能合約;區塊鏈;去中心化;因果規則探勘;smart contract;blockchain;decentralized;causal rule mining
Date:	2022-07-07
Issue Date:	2022-10-04 11:57:58 (UTC+8)
Publisher:	國立中央大學
Abstract:	本研究選擇去中心化的區塊鏈作為資料分享平台的系統架構，以區塊鏈和智能合約作為管理、協助和監督的角色，利用其固有的優勢，以不可更改的特性，可以追溯、監督平台上的所有動作，降低各參與組織間所需的信任門檻。本研究以較關聯規則探勘(association rule mining)更有目的性的因果規則探勘(causal rule mining)來顯現各組織間多面向和多維度的複雜關係。透過國立中央大學校務資料，來模擬校內組織間資料共享，並以欄位屬性多寡，區分為大規模、中規模和小規模的資料集。另外也透過美國職業籃球的資料，嘗試在不同的領域進行模擬，將這些規模複雜度、廣度有差距的資料進行因果規則分析比較，顯現在有效資料交流過後的資料環境下，結果有何差異。至於資料交流最在意的隱私保護，本研究會設計一個平台方作為管理資料的分享、匯集和分析的樞紐，各方參與分享的組織並不會直接接觸到其他組織的資料，以降低個人或各組織隱私外洩的可能，希望透過這樣的研究來推動整合社會上各領域整體資料的利用價值。最後根據本研究的實驗結果也顯現，將不同組織的資料進行合併，使有目的性的探勘資料集更豐富;這將使分析結果更具意義，奠定更確實的決策支援基礎。;Our research chooses a decentralized blockchain as the system architecture of the data sharing platform, using blockchain and smart contracts as management, assistance and supervision roles, to take advantage of their inherent advantages to trace all the data on the platform with its immutable characteristics to reduce the level of trust required between participating organizations. Our study uses causal rule mining which is more purposeful than association rule mining, to reveal the multi-faceted and multi-dimensional complex relationships among organizations. Through the data of National Central University, we simulated the data exchange among organizations in NCU, our study decided to distinguish large-scale, medium-scale and small-scale data sets by the number of column attributes. In addition, we also utilize the data from NBA and NCAA to simulate apply in different fields. As for privacy protection, our research will design a platform as a hub for managing data sharing, aggregation, and analysis, and the organizations involved in sharing will not directly participate in the data of other organizations to reduce the possibility of privacy leakage for individuals or organizations. It is hoped that such a study will promote and integrate the value of data utilization in various fields of society. Finally, the experiment showed that combining data from different organizations enriches the purposeful exploration data set; this will make the analysis results more meaningful and provide a more solid basis for decision support.
Appears in Collections:	[Graduate Institute of Computer Science and Information Engineering] Electronic Thesis & Dissertation

Files in This Item:

File	Description	Size	Format
index.html		0Kb	HTML	122	View/Open

社群 sharing

Loading...