臺灣博碩士論文加值系統

當全世界因為人工智慧而感到不可思議的同時，台灣面對人口迅速老化和低出生率的問題。隨著人口老齡化，照顧老年人健康的重要性是不言而喻的。台灣作為世界上其一擁有世界級的醫療水準的國家及其數一數二最完善的醫療保健系統及健保資料庫，其人才的培育也是在過去幾年在台灣政府對高等教育和研究的鼓勵下而建立了一個多元化的科技人才庫。台灣的醫療產業具有巨大的潛力，特別當人工智慧、機器學習與醫療產業相結合，這將台灣轉變升級成為人工智慧醫療保健第一大國。目前最讓人期待、興奮和鼓舞人心的智慧健康產業話題是精準醫療。其利用科技技術的進步和機器學習模型達到合適個人而非一體適用的醫療方案，從而改善患者治療效果並節省成本。面對台灣目前過時且嚴格的法律監管，法律成為了邁向精準醫療最大的障礙，因此當務之急是修法以促進智慧健康產業的發展。

人工智慧行動計劃能夠顯示一個國家對智慧醫療產業發展的態度和支持。唯有政府的實質支持和智慧健康產業的大家共同努力方能創造出台灣的明天。政府的資金補助固然重要，但其友善的法規環境也是留住人才和促進創新的重要因素之一。台灣目前的個人資料保護法和國民健康保險研究數據庫（NHIRD）的龐大醫療保健數據是產業目前討論最大的議題。邁向精準醫學和最大化人民健康福祉的同時，如何在大數據的收集與個人資料保護之間取得平衡讓人工智慧在台灣快速成長成了重要議題。台灣目前積極爭取GDPR的適足性認定，因此，根據台灣現行的個人資料保護法（PIPA），為未來修正個人資料保護法上進行分析和提供建議。台灣AI生態圈的建立是也是本文其中的討論策略方案。

While the world is amazed by AI, Taiwan is bewildered on critical nation issues which rapidly ageing population and low birth rates. With an ageing population, the significance of taking care of the old is plainly obvious. Howbeit, what being proud of is, Taiwan has earned recognition as one of the world’s best health care systems and also numerous of hospital has been ranked as top hospitals in the world. Furthermore, with the encouragement from Taiwan government on higher education and research in the past several years, Taiwan has trained a diverse pool of talents in scientific and technical fields. The smart healthcare industry in Taiwan is of great potential especially in combination with AI machine learning capabilities, and thus it has transformed Taiwan into the AI medical healthcare invention. The most exciting and encouraging topic of today among the peers in smart healthcare industry will be the AI in precision medicine. With the assistance of machine learning system and analytic skills, it moves us from one-size-fits-all healthcare to personalized, data-driven treatment that improved patient outcomes and cost saving. Nonetheless, less startups and companies in Taiwan have been able to start capitalizing on the precision medicine, owing to the biggest barriers in which the uncertain legal and regulatory framework and lack of access to relevant data. Facing with the current outdated, strict and rigid regulation in Taiwan, an industry-friendly legislation is urgently needed in order to keep pace with the AI trend and promote industrial development. An industry-friendly policy is the condition to foster domestic demand and industrial development.

Nation’s AI action plan is able to show attitude and support of a country on AI healthcare development. Nation’s AI action plan is highly important as with the support from the government and the effort with peers industry, AI healthcare industry in Taiwan definitely could have significant development. Funding is indeed important for industry development. A creation of friendly legislative environment to keep talents and promote startup for innovation is important too. As known, AI having inseparable relationship with big data. In order to promote and maximize the welfare of a nation thru AI precision medicine, healthcare datasets of nation is then a key issue to achieve it. Data analysis and data mining encompasses personal data protection question. A balance to strike between personal privacy during data mining and public interest in order to adapt with the rapid growth of AI. Therefore, challenges that arise and the opportunities on Taiwan’s current privacy law on personal data protection and the healthcare data on National Health Insurance Research Database (NHIRD) will then be one of the important issues in Taiwan’s AI healthcare development. With the current strict and rigorous protection on privacy and personal data protection in Taiwan, it seems blocking the movement of AI development towards the precision medicine. Since Taiwan is on the track towards GDPR adequacy decision, it is therefore, based on Taiwan’s current Personal Information Protection Act (PIPA), analysis made for the suggestion for PIPA amendment in future. Building of AI ecosystem is one of the strategies discussed in this research.

Chinese Abstract i
English Abstract iii
Acknowledgments v
Table of Contents vi
List of Figures viii
Chapter 1 Research Introduction 1
1.1 Research Background and Objectives 1
1.2 Methodology, Scope and Limitation 3
1.2.1 Methodology 3
1.2.2 Scope 4
1.2.3 Limitation 5
1.3 Research Framework 6
Chapter 2 AI Evolution and Development on Smart Healthcare Industry 8
2.1 AI Resurgence 8
2.1.1 What is Artificial Intelligence? 9
2.1.2 AI Road to Enlightenment and its Waves 24
2.1.2.1 The Heel of Achilles: 1st Wave of AI (1956~1974) 24
2.1.2.2 Reignited Passion: 2nd Wave of AI (1980~1987) 25
2.1.2.3 Disillusionment of Hope: AI Winters 25
2.1.2.4 Big Stir: 3rd Wave of AI (2010~Today) 26
2.1.2.5 Today, Future: 4th Wave of AI 27
2.2 AI Industrial Application, Current Trends and Market Perspectives 30
2.3 Taiwan’s Position in 4th Wave of AI 37
2.3.1 Why Taiwan in Smart Healthcare 37
2.4 Analysis 42
2.5 Summary 44
Chapter 3 Development Strategies on Healthcare and National Action Plan 49
3.1 National Artificial Intelligence Strategies 49
3.1.1 Canada 51
3.1.1.1 Pan-Canadian Artificial Intelligence Strategy 53
3.1.1.2 Pan-Canadian Artificial Intelligence Strategy in Healthcare 54
3.1.2 Taiwan 57
3.1.2.1 Taiwan AI Action Plan 60
3.1.2.2 Taiwan AI Action Plan in Healthcare 65
3.2 Analysis, Discussion and Suggestion 67
3.2.1 Suggestion 68
3.3 Summary 70
Chapter 4 Smart Health and Data Privacy 73
4.1 Global Healthcare Industry 73
4.1.1 Antagonistic Relation of AI and Personal Data 73
4.2 Smart Health and GDPR 75
4.2.1 General Data Protection Act (GDPR) 75
4.2.2 Smart Health under GDPR 86
4.3 Reference of Data Protection 87
4.3.1 Canada 87
4.3.1.1 Canada: Data Storage System 88
4.3.1.2 Canada: Data Protection 92
4.3.1.3 Comparison Between the GDPR and Canada’s Privacy Laws 94
4.3.2 Taiwan 100
4.3.2.1 Taiwan: Data Storage System 103
4.3.2.2 Comparison Between the GDPR and Taiwan’s Privacy Laws 105
4.3.2.3 Attitude of GDPR and Taiwan’s Privacy Law on Research Field 111
4.4 Analysis, Discussion and Suggestion 114
4.4.1 Suggestion 116
4.5 Summary 127
Chapter 5 Conclusion and Suggestion 132
5.1 Conclusion 132
5.2 Suggestion 135
5.2.1 Further Research in Future 148
References 150
Slides Share 153

1.John McCarthy et al., A Proposal for the Dartmouth Summer Research Project on Artificial Intelligence, August 31, 1955, 27 AI MAGAZINE, 2006.
2.Jeremy Horwitz, GOOGLE ASSISTANT WINS AI IQ TEST; SIRI BEATS ALEXA AND CORTANA VENTURE BEAT (2018), https://venturebeat.com/2018/07/25/google-assistant-wins-ai-iq-test-siri-beats-alexa-and-cortana/ (last visited Dec 1, 2018).
3.Kristin Houser, DEEPMIND HAVE CREATED AN IQ TEST FOR AI, AND IT DIDN'T DO TOO WELL, WORLD ECONOMIC FORUM (2018), https://www.weforum.org/agenda/2018/07/deepmind-created-a-test-to-measure-an-ai-s-ability-to-reason (last visited Dec 1, 2018).
4.Measuring abstract reasoning in neural networks, DEEPMIND, https://deepmind.com/blog/measuring-abstract-reasoning/ (last visited Dec 1, 2018).
5.Utpal Chakraborty, WE KNOW VERY LITTLE ABOUT HUMAN BRAIN: ARTIFICIAL NEURAL NETWORK VS. MEDIUM CORPORATION (2018), https://medium.com/@utpal_bob/we-know-very-little-about-human-brain-artificial-neural-network-vs-66e4ac20a508 (last visited May 9, 2019).
6.Gaetano Licata, 2015, Are Neural Networks Imitations of Mind?, J COMPUT SCI SYST BIOL 8:124-126.
7.FU KAI LEE & GANG YONG WANG, AI IS HERE, TAIPEI: CHEERS, 2017, PG 15-201.
8.Klaus Schwab, THE FOURTH INDUSTRIAL REVOLUTION: WHAT IT MEANS AND HOW TO RESPOND, WORLD ECONOMIC FORUM (2016), https://www.weforum.org/agenda/2016/01/the-fourth-industrial-revolution-what-it-means-and-how-to-respond/ (last visited May 9, 2019).
9.Priya Dwivedi, TRACKING A SELF-DRIVING CAR WITH HIGH PRECISION, TOWARDS DATA SCIENCE (2017), https://towardsdatascience.com/helping-a-self-driving-car-localize-itself-88705f419e4a (last visited May 9, 2019).
10.AI Industry New Era, Big Breakthrough on Medical Education, DA WIN DINING (2018), https://www.youtube.com/watch?v=h6aEhJKd4Q8 (last visited May 9, 2019).
11.Gary Marcus, Researchers Need New Ways to Distinguish Artificial Intelligence From the Natural Kind, SCIENTIFIC AMERICAN, 2017, at 59–63.
12.Artificial Intelligence Deals Tracker: 5K Deals Across 25 Industries In One Heatmap, CB INSIGHTS RESEARCH (2018), https://www.cbinsights.com/research/artificial-intelligence-investment-heatmap-expert-intelligence/ (last visited May 9, 2019).
13.David Lee, AI ENTREPRENEURIAL INNOVATION FOR THE HAPPINESS OF GLOBAL SOCIETY (2018).
14.R. Colin Johnson, MICROSOFT, GOOGLE BEAT HUMANS AT IMAGE RECOGNITION, EE TIMES (2015), https://www.eetimes.com/document.asp?doc_id=1325712#.
15.Priya Dwivedi, TRACKING A SELF-DRIVING CAR WITH HIGH PRECISION, TOWARDS DATA SCIENCE (2017), https://towardsdatascience.com/helping-a-self-driving-car-localize-itself-88705f419e4a (last visited May 9, 2019).
16.Jonathan Vanian, IBM BOUGHT THE WEATHER COMPANY BECAUSE WEATHER AFFECTS NEARLY EVERYTHING, FORTUNE (2015), http://fortune.com/2015/10/28/ibm-weather-company-acquisition-data/ (last visited May 9, 2019).
17.Don Weinland, GLOBAL M&A ACTIVITY HITS NEW HIGH, FINANCIAL TIMES (2018), https://www.ft.com/content/b7e67ba4-c28f-11e8-95b1-d36dfef1b89a (last visited May 9, 2019).
18.Gartner Identifies Five Emerging Technology Trends That Will Blur the Lines Between Human and Machine, Gartner (2018), https://www.gartner.com/en/newsroom/press-releases/2018-08-20-gartner-identifies-five-emerging-technology-trends-that-will-blur-the-lines-between-human-and-machine (last visited May 9, 2019).
19.Global Talent 2021, OXFORD ECONOMICS, https://www.oxfordeconomics.com/.
20.Reenita Das, TOP 8 HEALTHCARE PREDICTIONS FOR 2019, FORBES (2018), https://www.forbes.com/sites/reenitadas/2018/11/13/top-8-healthcare-predictions-for-2019/#434093fd700e (last visited May 9, 2019).
21.Laura Donnelly, DOCTORS ARE SO TIRED FROM LONG SHIFTS THAT THEY 'BEHAVE LIKE THEY ARE DRUNK', THE TELEGRAPH(2018), https://www.telegraph.co.uk/news/2018/06/25/doctors-tired-long-shifts-behave-like-drunk/ (last visited May 9, 2019).
22.AI report from McKinsey Global Institute: AI to spur economic growth, INFORMATION AGE (2018), https://www.information-age.com/mckinsey-global-institute-123474592/ (last visited May 9, 2019).
23.Forget deep learning, unsupervised deep learning is the future, Information Age (2018), https://www.information-age.com/unsupervised-deep-learning-123477727/ (last visited May 9, 2019).
24.William Watson, HEALTH CARE'S FREE IN CANADA, RIGHT? WRONG, AND YOU'RE PAYING A LOT MORE THAN YOU THINK, FINANCIAL POST(2017), https://business.financialpost.com/opinion/health-cares-free-in-canada-right-wrong-and-youre-paying-a-lot-more-than-you-think (last visited May 14, 2019).
25.Waiting Your Turn: Wait Times for Health Care in Canada, 2018 Report, FRASER INSTITUTE (2018), https://www.fraserinstitute.org/studies/waiting-your-turn-wait-times-for-health-care-in-canada-2018 (last visited May 14, 2019).
26.Steven Hoffman, RESEARCH INNOVATION THE GREAT PROMISE OF ARTIFICIAL INTELLIGENCE FOR PUBLIC HEALTH COMMENTS, http://research.info.yorku.ca/2018/04/the-great-promise-of-artificial-intelligence-for-public-health/ (last visited May 14, 2019).
27.Richard Birtwhistle, Marshall Godwin, Jannet Ann Leggett and Ken Martin, Linking Health Databases for Research, Canadian Family Physician April 2015, vol. 61 issue (4) pg.382
28.Joanne Shih, Big Data Era: Nation Bio-Database Upgrading Bio-Medical Standard, Gene Online News (2016), https://geneonline.news/index.php/2016/04/28/bigdataandbiobank/ (last visited May 13, 2019).
29.A. Gittens et al., UNDERSTANDING THE GDPR: A COMPARISON BETWEEN THE GDPR, PIPEDA AND PIPA, LEXOLOGY (2018), https://www.lexology.com/library/detail.aspx?g=64796406-67e5-49e2-94b7-78a494f97311 (last visited May 13, 2019).
30.REGULATION (EC) No 1338/2008 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL of 16 December 2008 on Community statistics on public health and health and safety at work, Official Journal of the European Union (2008), https://eur-lex.europa.eu/legal-content/EN/TXT/PDF/?uri=CELEX:32008R1338&rid=2 (last visited May 13, 2019).