基於非侵入式手機使用者識別機制即時檢測使用者操作行為收集其建模資料的方法;Using Active Learning to Collect User’s Behavior for Training Model. Base on Non-intrusive Smartphone Authentication

NCUIR > College of Electrical Engineering & Computer Science > Graduate Institute of Computer Science and Information Engineering > Electronic Thesis & Dissertation > Item 987654321/68738

Please use this identifier to cite or link to this item: http://ir.lib.ncu.edu.tw/handle/987654321/68738

Title:	基於非侵入式手機使用者識別機制即時檢測使用者操作行為收集其建模資料的方法;Using Active Learning to Collect User’s Behavior for Training Model. Base on Non-intrusive Smartphone Authentication
Authors:	陳懿婷;Chen,Yi-Ting
Contributors:	資訊工程學系
Keywords:	非侵入式識別機制;使用者識別;主動學習;支持向量機;non-intrusive authentication mechanism;user authentication;active learning;support vector machine
Date:	2015-07-21
Issue Date:	2015-09-23 14:22:50 (UTC+8)
Publisher:	國立中央大學
Abstract:	隨著科技進步，智慧型手機可以經由網路連結，達到收發電子信箱、線上交易(網路銀行、電子商務)、社群軟體等功能，無形中使智慧型手機內部資料成為有心人士覬覦的目標。目前智慧型手機識別機制有侵入式與非侵入式兩種。傳統的驗證機制（密碼鎖、圖形鎖）屬於侵入式識別機制(一次驗證)。非侵入式識別機制則不需要驗證介面，而是從背景收集使用者行為進行驗證。目前文獻上已有提出非侵入式識別機制數種研究，但皆未考慮訓練樣本選擇與數目，在實際應用上若不考慮此點將耗費使用者許多時間在提供訓練樣本上。本研究提出即時檢測收集方法利用主動學習搭配支持向量機選擇訓練樣本，在識別效果接受範圍內以少量的訓練樣本建構非侵入式識別機制。首先本研究提出兩階段資料收集方法，第一階段偵測個人所需的行為樣本情境，第二階段則針對第一階段所分析出的情境加以收集訓練樣本並建模。最後將本論文提出的即時檢測收集方法與批次性方法比較，實驗結果則有一半以上使用者能透過本方法在識別效果不受影響下大幅減少訓練樣本。 ;In order to protect the data within the smartphone, intrusive and non-intrusive user authentication mechanisms were developed. Traditional authentication mechanisms like number lock and pattern lock are intrusive user authentication mechanism. Non-intrusive user authentication mechanism doesn’t require any user interface, but collect user’s behavior in the background and authenticate it. Several non-intrusive authentication mechanisms were proposed, but all of them don’t care about the selection of training samples. Actually user to provide the training samples can be very time-consuming. This study proposes a method to collect real-time detection with the use of active learning support vector machine choose training samples to identify the effect of the acceptable range in a small amount of training samples construction of non-invasive identification mechanism. First, for this study, we propose two stages collection methodology of behavior. The first stage of the collection is to detect which behavior scenario need. The second stage of the collection is to for the first phase of the analysis of the situation and to collect training data modeling. Finally, this study presents a method to collect real-time detection (active learning) compared with batch learning, results are more than half of users to significantly reduce the training sample to achieve a good recognition results through this method.
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	335	View/Open

社群 sharing

Loading...