Publications

Detailed Information

Designing Flexible Self-Tracking Technologies for Enhancing In Situ Data Collection Capability : 현장 데이터 수집 능력을 확장하기 위한 자유도 높은 셀프 트래킹 기술의 디자인

DC Field Value Language
dc.contributor.advisor서진욱-
dc.contributor.author김영호-
dc.date.accessioned2019-05-07T05:35:27Z-
dc.date.available2019-05-07T05:35:27Z-
dc.date.issued2019-02-
dc.identifier.other000000154408-
dc.identifier.urihttps://hdl.handle.net/10371/151973-
dc.description학위논문 (박사)-- 서울대학교 대학원 : 공과대학 컴퓨터공학부, 2019. 2. 서진욱.-
dc.description.abstractCollecting and tracking data in everyday contexts is a common practice for both individual self-trackers and researchers. The increase in wearable and mobile technologies for self-tracking encourages people to gain personal insights from the data about themselves. Also, researchers exploit self-tracking to gather data in situ or to foster behavioral change.



Despite a diverse set of available tracking tools, however, it is still challenging to find ones that suit unique tracking needs, preferences, and commitments. Individual self-tracking practices are constrained by the tracking tools' initial design, because it is difficult to modify, extend, or mash up existing tools. Limited tool support also impedes researchers' efforts to conduct in situ data collection studies. Many researchers still build their own study instruments due to the mismatch between their research goals and the capabilities of existing toolkits.



The goal of this dissertation is to design flexible self-tracking technologies that are generative and adaptive to cover diverse tracking contexts, ranging from personal tracking to research contexts. Specifically, this dissertation proposes OmniTrack, a flexible self-tracking approach leveraging a semi-automated tracking concept that combines manual and automated tracking methods to generate an arbitrary tracker design.

OmniTrack was implemented as a mobile app for individuals. The OmniTrack app enables self-trackers to construct their own trackers and customize tracking items to meet their individual needs. A usability study and a field development study were conducted with the goal of assessing how people adopt and adapt OmniTrack to fulfill their needs. The studies revealed that participants actively used OmniTrack to create, revise, and appropriate trackers, ranging from a simple mood tracker to a sophisticated daily activity tracker with multiple fields.



Furthermore, OmniTrack was extended to cover research contexts that enclose manifold personal tracking contexts. As part of the research, this dissertation presents OmniTrack Research Kit, a research platform that allows researchers without programming expertise to configure and conduct in situ data collection studies by deploying the OmniTrack app on participants' smartphones. A case study in deploying the research kit for conducting a diary study demonstrated how OmniTrack Research Kit could support researchers who manage study participants' self-tracking process.



This work makes artifacts contributions to the fields of human-computer interaction and ubiquitous computing, as well as expanding empirical understanding of how flexible self-tracking tools can enhance the practices of individual self-trackers and researchers. Moreover, this dissertation discusses design challenges for flexible self-tracking technologies, opportunities for further improving the proposed systems, and future research agenda for reaching the audiences not covered in this research.
-
dc.description.abstract일상의 맥락에서 데이터를 모으는 활동인 셀프 트래킹(self-tracking)은 개인과 연구의 영역에서 활발히 활용되고 있다. 웨어러블 디바이스와 모바일 기술의 발달로 인해 사람들은 각자의 삶에 대해 말해주는 데이터를 더 쉽게 수집하고, 통찰할 수 있게 되었다. 또한, 연구자들은 현장(in situ) 데이터를 수집하거나 사람들에게 행동 변화를 일으키는 데에 셀프 트래킹을 활용한다.



비록 셀프 트래킹을 위한 다양한 도구들이 존재하지만, 트래킹에 대해 다양화된 요구와 취향을 완벽히 충족하는 것들을 찾는 것은 쉽지 않다. 대부분의 셀프 트래킹 도구는 이미 설계된 부분을 수정하거나 확장하기에 제한적이다. 그렇기 때문에 사람들의 셀프 트래킹에 대한 자유도는 기존 도구들의 디자인 공간에 의해 제약을 받을 수밖에 없다. 마찬가지로, 현장 데이터를 수집하는 연구자들도 이러한 도구의 한계로 인해 여러 문제에 봉착한다. 연구자들이 데이터를 통해 답하고자 하는 연구 질문(research question)은 분야가 발전할수록 세분되고, 치밀해지기 때문에 이를 위해서는 복잡하고 고유한 실험 설계가 필요하다. 하지만 현존하는 연구용 셀프 트래킹 플랫폼들은 이에 부합하는 자유도를 발휘하지 못한다. 이러한 간극으로 인해 많은 연구자들이 각자의 현장 데이터 수집 연구에 필요한 디지털 도구들을 직접 구현하고 있다.



본 연구의 목표는 자유도 높은---연구적 맥락과 개인적 맥락을 아우르는 다양한 상황에 활용할 수 있는---셀프 트래킹 기술을 디자인하는 것이다. 이를 위해 본고에서는 옴니트랙(OmniTrack)이라는 디자인 접근법을 제안한다. 옴니트랙은 자유도 높은 셀프 트래킹을 위한 방법론이며, 반자동 트래킹(semi-automated tracking)이라는 컨셉을 바탕으로 수동 방식과 자동 방식의 조합을 통해 임의의 트래커를 표현할 수 있다.



먼저 옴니트랙을 개인을 위한 모바일 앱 형태로 구현하였다. 옴니트랙 앱은 개개인이 자신의 트래킹 니즈에 맞는 트래커를 커스터마이징하여 활용할 수 있도록 구성되어 있다. 본고에서는 사람들이 어떻게 옴니트랙을 자신의 니즈에 맞게 활용하는지 알아보고자 사용성 테스트(usability testing)와 필드 배포 연구(field deployment study)를 수행하였다. 참가자들은 옴니트랙을 활발히 이용해 다양한 디자인의 트래커—아주 단순한 감정 트래커부터 여러 개의 필드를 가진 복잡한 일일 활동 트래커까지—들을 생성하고, 수정하고, 활용하였다.



다음으로, 옴니트랙을 현장 데이터 수집 연구에 활용할 수 있도록 연구 플랫폼 형태의 '옴니트랙 리서치 킷(OmniTrack Research Kit)'으로 확장하였다. 옴니트랙 리서치 킷은 연구자들이 프로그래밍 언어 없이 원하는 실험을 설계하고 옴니트랙 앱을 참가자들의 스마트폰으로 배포할 수 있도록 디자인되었다. 그리고 옴니트랙 리서치 킷을 이용해 일지기록 연구(diary study)를 직접 수행하였고, 이를 통해 옴니트랙 접근법이 어떻게 연구자들의 연구 목적을 이루는 데에 도움을 줄 수 있는지 직접 확인하였다.



본 연구는 휴먼-컴퓨터 인터랙션(Human-Computer Interaction) 및 유비쿼터스 컴퓨팅(Ubiquitous Computing) 분야에 기술적 산출물로써 기여하며, 자유도 높은 셀프 트래킹 도구가 어떻게 개인과 연구자들을 도울 수 있는지 실증적인 이해를 증진한다. 또한, 자유도 높은 셀프트래킹 기술에 대한 디자인적 난제, 연구에서 제시한 시스템에 대한 개선방안, 마지막으로 본 연구에서 다루지 못한 다른 집단을 지원하기 위한 향후 연구 논제에 대하여 논의한다.
-
dc.description.tableofcontentsAbstract

CHAPTER 1. Introduction

1.1 Background and Motivation

1.2 Research Questions and Approaches

1.2.1 Designing a Flexible Self-Tracking Approach Leveraging Semiautomated

Tracking

1.2.2 Design and Evaluation of OmniTrack in Individual Tracking Contexts

1.2.3 Designing a Research Platform for In Situ Data Collection Studies

Leveraging OmniTrack

1.2.4 A Case Study of Conducting an In Situ Data Collection Study

using the Research Platform

1.3 Contributions

1.4 Structure of this Dissertation



CHAPTER 2. Related Work

2.1 Background on Self-Tracking

2.1.1 Self-Tracking in Personal Tracking Contexts

2.1.2 Utilization of Self-Tracking in Other Contexts

2.2 Barriers Caused by Limited Tool Support

2.2.1 Limited Tools and Siloed Data in Personal Tracking

2.2.2 Challenges of the Instrumentation for In Situ Data Collection

2.3 Flexible Self-Tracking Approaches

2.3.1 Appropriation of Generic Tools

2.3.2 Universal Tracking Systems for Individuals

2.3.3 Research Frameworks for In Situ Data Collection

2.4 Grounding Design Approach: Semi-Automated Tracking

2.5 Summary of Related Work



CHAPTER3

DesigningOmniTrack: a Flexible Self-Tracking Approach

3.1 Design Goals and Rationales

3.2 System Design and User Interfaces

3.2.1 Trackers: Enabling Flexible Data Inputs

3.2.2 Services: Integrating External Trackers and Other Services

3.2.3 Triggers: Retrieving Values Automatically

3.2.4 Streamlining Tracking and Lowering the User Burden

3.2.5 Visualization and Feedback

3.3 OmniTrack Use Cases

3.3.1 Tracker 1: Beer Tracker

3.3.2 Tracker 2: SleepTight++

3.3.3 Tracker 3: Comparison of Automated Trackers

3.4 Summary



CHAPTER 4. Understanding HowIndividuals Adopt and Adapt OmniTrack

4.1 Usability Study

4.1.1 Participants

4.1.2 Procedure and Study Setup

4.1.3 Tasks

4.1.4 Results and Discussion

4.1.5 Improvements A_er the Usability Study

4.2 Field Deployment Study

4.2.1 Study Setup

4.2.2 Participants

4.2.3 Data Analysis and Results

4.2.4 Reflections on the Deployment Study

4.3 Discussion

4.3.1 Expanding the Design Space for Self-Tracking

4.3.2 Leveraging Other Building Blocks of Self-Tracking

4.3.3 Sharing Trackers with Other People

4.3.4 Studying with a Broader Audience

4.4 Summary



CHAPTER 5. Extending OmniTrack for Supporting In Situ Data Collection

Studies

5.1 Design Space of Study Instrumentation for In-Situ Data Collection

5.1.1 Experiment-Level Dimensions

5.1.2 Condition-Level Dimensions

5.1.3 Tracker-Level Dimensions

5.1.4 Reminder/Trigger-Level Dimensions

5.1.5 Extending OmniTrack to Cover the Design Space

5.2 Design Goals and Rationales

5.3 System Design and User Interfaces

5.3.1 Experiment Management and Collaboration

5.3.2 Experiment-level Configurations

5.3.3 A Participants Protocol for Joining the Experiment

5.3.4 Implementation

5.4 Replicated Study Examples

5.4.1 Example A: Revisiting the Deployment Study of OmniTrack

5.4.2 Example B: Exploring the Clinical Applicability of a Mobile Food

Logger

5.4.3 Example C: Understanding the Effect of Cues and Positive Reinforcement

on Habit Formation

5.4.4 Example D: Collecting Stress and Activity Data for Building a

Prediction Model

5.5 Discussion

5.5.1 Supporting Multiphase Experimental Design

5.5.2 Serving as Testbeds for Self-Tracking Interventions

5.5.3 Exploiting the Interaction Logs

5.6 Summary



CHAPTER 6. Using the OmniTrack Research Kit: A Case Study

6.1 Study Background and Motivation

6.2 OmniTrack Configuration for Study Instruments

6.3 Participants

6.4 Study Procedure

6.5 Dataset and Analysis

6.6 Study Result

6.6.1 Diary Entries

6.6.2 Aspects of Productivity Evaluation

6.6.3 Productive Activities

6.7 Experimenter Experience of OmniTrack

6.8 Participant Experience of OmniTrack

6.9 Implications

6.9.1 Visualization Support for Progressive, Preliminary Analysis of

Collected Data

6.9.2 Inspection to Prevent Misconfiguration

6.9.3 Providing More Alternative Methods to Capture Data

6.10 Summary



CHAPTER 7. Discussion

7.1 Lessons Learned

7.2 Design Challenges and Implications

7.2.1 Making the Flexibility Learnable

7.2.2 Additive vs. Subtractive Design for Flexibility

7.3 Future Opportunities for Improvement

7.3.1 Utilizing External Information and Contexts

7.3.2 Providing Flexible Visual Feedback

7.4 Expanding Audiences of OmniTrack

7.4.1 Supporting Clinical Contexts

7.4.2 Supporting Self-Experimenters

7.5 Limitations



CHAPTER 8. Conclusion

8.1 Summary of the Approaches

8.2 Summary of Contributions

8.2.1 Artifact Contributions

8.2.2 Empirical Research Contributions

8.3 Future Work

8.3.1 Understanding the Long-term E_ect of OmniTrack

8.3.2 Utilizing External Information and Contexts

8.3.3 Extending the Input Modality to Lower the Capture Burden

8.3.4 Customizable Visual Feedback

8.3.5 Community-Driven Tracker Sharing

8.3.6 Supporting Multiphase Study Design

8.4 Final Remarks



APPENDIX A. Study Material for Evaluations of the OmniTrack App

A.1 Task Instructions for Usability Study

A.2 The SUS (System Usability Scale) Questionnaire

A.3 Screening Questionnaire for Deployment Study

A.4 Exit Interview Guide for Deployment Study

A.5 Deployment Participant Information



APPENDIX B

Study Material for Productivity Diary Study

B.1 Recruitment Screening Questionnaire

B.2 Exit Interview Guide

Abstract (Korean)
-
dc.language.isoeng-
dc.publisher서울대학교 대학원-
dc.subject.ddc621.39-
dc.titleDesigning Flexible Self-Tracking Technologies for Enhancing In Situ Data Collection Capability-
dc.title.alternative현장 데이터 수집 능력을 확장하기 위한 자유도 높은 셀프 트래킹 기술의 디자인-
dc.typeThesis-
dc.typeDissertation-
dc.contributor.AlternativeAuthor김영호-
dc.description.degreeDoctor-
dc.contributor.affiliation공과대학 컴퓨터공학부-
dc.date.awarded2019-02-
dc.identifier.uciI804:11032-000000154408-
dc.identifier.holdings000000000026▲000000000039▲000000154408▲-
Appears in Collections:
Files in This Item:

Altmetrics

Item View & Download Count

  • mendeley

Items in S-Space are protected by copyright, with all rights reserved, unless otherwise indicated.

Share