Data based digital developmen policy and strategy ChangHee Yun

Data based digital developmen policy and strategy ChangHee Yun Intelligent Information Technology Team

INDEX Ⅰ. Ⅱ. Ⅲ. Ⅳ . V. Data-Critical factor of the 4th Industrial Revolution Utilize data in Korea Strategy for building data Strategy for building AI data Social change by data 2

I. Data, Critical factor of the 4th Revolution Industrial 3

1 Data-Critical factor of the 4th Industrial Revolut Innovation Growth for Human by utilizing data Data for Growth and Wellbeing capital AI (IBM watson) autonomous vehicle AlphaG o Smart City 4th Industrial Data Revolution Revolution Smart Factory

2 Data-Critical factor of the 4th Industrial Revolut Data –Crude Oil of the 4th Industrial Revolution Data Connect & Fusion : Creating new value in intelligence information society - Combining and analyzing different types of data is likely to discover new values ex) Quarantine information data roaming data of telecommunication Corp. block infectious disease from overseas Call data at late midnight Taxi riding&getting off data Efficiency of late night bus route - Activation of data industry needed for generation and distribution of valueadded data

3 Data-Critical factor of the 4th Industrial Revolut New application of data emerged as the core infrastructure of intelligent information society - Realization of intelligent information tech needs for various training date s - IoT, smart devices make the data of people and things increase * Global IoT device sales prospeted to reach 25 billion units by 2020 (Gartner, 14) * The global digital scale is expected to increase tenfold over seven years (IDC, 15) * Data generated over the past two years accounts for 90% of the world's data (IBM, 12) “90% of the data in the world today has been created in the last two years alone”(IBM, ‘12)

II. Utilize data in Korea 7

1 Data utilization in Korea-ICBM(Data, Clou Data has been utilized centered on each institution and company in public and private sector Big data four issues Smart administratio n Night Bus KT Corp. Lack of useful data Public Privat Competitively Establish e introducedby centralgovernment andmunicipal government Restriction, use of personal information Big Data Center APT Maintenance Fee Traditional Village KB Financial group 케이웨더 CJ Corp. Inadequate Inadequatedata datastandardization standardization Lack Lack of of big big data data experts experts operated 92 big data centers in telecommunication, finance, manufacturing, public, university(as - Impossible to collect big data in one place due to large capacity, real time, and personal information Need to build a big data network that links key data production and collection agencies

.2 Utilize data in Korea-ICBM(Data, Cloud) Data production and opening rate are high - The mobile traffic per capita is 963MB monthly, which is 10 times the world's usage rate (Cisco, 15) - Public data is opened about 210,000 as of October ‘16 89 M 20,679 963 M 10Times 3.9Times 5,272 World average Korea Mobile traffic per capita End of 2013 End of 2016 OECD 공공데이터 개방지수 1 위 (OECD, '15) Data utilization is insuffi cient. - Big data utilization is one-third of world(29%) 29 % World average 1/3 (Tech Pro Research, 15) 9.6 % Korea Big data utilization rate Cloud traffi c rate : OECD(86%), Korea(1.35%) 9

.3 Utilize data in Korea Lack of active data due to Legal system privacy regulations - Regulation of pre-agreement method for collection and utilization of personal informat * Korea is the highest personal information regulation in Asia(British law firm ‘Hogan Lobel - Negative recognition of the use of personal information Low utilization of open public data - Difficulty to use due to different format between open data - Lower quality of free public data released 10

III. strategy for building data 11

1.1 Strengths in Korea Securing the world's best ICT infrastructure through national ICT effor - Achieved 1st rank in ICT development index for two years in a row('15 '16, ITU) Accessibility and availability are highly evaluated in the field of public data openness - 1st rank in Public Data Opening Index ('17 .7, OECD 'Government White Paper’) A global testbed based on national character familiar with latest techno - An environment where the 4th industrial revolution technology can be expended actively Government-led data infrastructure to be based on intelligent information technology

.2 Data network diffusion strategy Data based leading future change in the field of society/economy and creating new industr Vision 전 Goal Key challenge Expanding data capital goods and strengthening 4th industrial ecosystem through building data network 1) Nationwide data center development 2) Data support center operation 3) Create data Industry ecosystem 4) Build data sharing and link environment 5) Data Cooperative Network Operation Telecommunication/Finance/Manufacture/Distribution Data Networkamong data expertise center (2018 2022) Region Public Innovation Innovation lLocalGov Service Analysis corp. Solution corp. R&D Health Medical Univ. Data produce r Energy Business Distributio n Finance Telecomm. Build data industry ecosystem Univ. Startup Institutio n Local Gov Data broker *ex) nationwide data capital expansion & distribution infra such as distribution network

.3 Build data industry infrastructure 우수 사례 공유확산 우수 사례 공유확산 Foster data startup Excavate new business Facilitate data distribution transaction Privat e Promote expertise Data process corp. Univ . Foster R&D Training experts Small business analysis consulting corp. Medium size solution corp. Analysis agency corp. 우수 사례 공유확산 Public Boost data open Support data foundation Resolution social issues Create market demand and secure industrial competitiveness through fostering expertise company such as data process, solution, analysis process etc.

Build data share and connect infrastruc Make data platform to share and connect public, private data DATA MAP News Credit card KT (Telecommu ) agriculture and fisheries GDSconsulting (Logistics) environment Kepco (Energy) Road Traffic Auth ority (Traffic) Kweather (Weather) Establish utilizing environment data map based floating population Media Logistics Cj olivenetworks (Food) Health insurance (Medical) NIA (Research) Connectmeta information Enrollmentforconnect structureamongdata Data support Center Connect data platform * LOD : Linked Open Data, Open standard (utilizing open data such asdata LOD, CSV etc.) * CSV : Coma Separated Value, text data to separate article based on comma Integrated management of open/ distribution data meta information Connect open data catalog(DCAT) (Open/distribution meta information automatic * DCAT: Data Catalog Vocabulary, web based data catalog update) interworking standard Operating Open data platform(CKAN) (Meta information standardization) Data expertise center Manufac, Distributi on Traffic, satety Telecomm,Common sense, social media * CKAN: Comprehensive Knowledge Archive Network, web based open data platform to be able to save/ distribution

.5 Operating data network cooperation Data cooperation network based promotion of excellence, sharing and cooperation busin Share best practices Data network council Share know-how such as data processing and business implementation examples through quarterly regular meetings (Data planning and policy advice) secretari (Promotion, event support, at etc.) Field Working Group (10 (Data standard, data link utilization cooperation etc.) divisions) Data linkage cooperation Collaborate on open / distribution data and business utilization data Finding a joint project Participate in data centers by sector Make opportunities for commercialization through business and agency cooperation business such as communication, card, transportation, logistics Communicatio n / Media . Manufacturin g/ Distribution Medical / Health Finding standard cooperation items Identify and cooperate with standards related to data opening / distribution and linkage utilization

VI. Strategy for building AI data 17

.1 4th Industrial Revolution and AI Connect Human·Things I oT share C loud Analysis B Ig data Device M obile “AI reanalysis” rtificial A I ntelligence Destructive innovation of existing industry Realize AI industry power

1.2 Vision and Goal Building the AI data infrastructure for realization of advanced AI industry - Make the world's best AI data infrastructure available to everyone by 2021 Vision Public data R&D Data Build national AI data infrastructure Goal AI inference training data set Refining ProcessingFusion Establish AI data distribution system 1. Knowledge base (Law, Tourism, Medical.) Industry AI Startup Private data Laboratory SW Corp. 2. Providing AI use environment 3. AI infra open platform AI R&D Hospital Data Corp. Exobrain DeepView MS Cortana ICT fusion project National R&D

1.3 Overview of building of AI Industry Infrastructure Project Create ‘AI Open Innovation Hub’ to provide one-stop support services for infrastructures needed for development and commercialization of Artificial Intelligence technologies. * Aggregation of government AI algorithms(open source/open API), machine learning data and large scale computing resources AI SW AI HW AI Algorithm CPU, GPU, Public AI SW (TensorFlow, Caffe) Memory AI Data Private Sector Data AI Data Laws, patents knowledge base Images specific to Korea Wikipedia SKT, KT Public Sector Data R&D Data Public Data Portal Data Research Center Standardization Human-care Robot Brain Computer Interface Material Information Personal Information De-identification

1.4 Korea strategy –AI data Build massive data sets that machines can learn Establishment of national data management system to enable AI machine learni - Public data is converted into a machine-readable format and opened Open in real time to systematically accumulate sensor data for IoT device production - Providing smart city's IoT data in connection with data platform Knowledge base(Ontology, LOD)

1.5 Korea strategy –AI data standard 50/60Hz 100/110/120/125/127/200/210/220/230/240 V 22

1.6 Purpose of Standard Knowledge base, which is data specialized for AI learning Securing universality, reusing, efficient linkage ‧ distribution The way to use and save AI data is ontology Most existing text-based AI learning data follows ontology method * Define standards for Class (Concept), Property (Attribute) and Relation Google Knowledge Graph Apple Siri Active Ontology Amazon Alexa Ontology

1.7 Korea AI standard plan Verify and establish the knowledge base standard for each field, the result of building a knowledge base designed and built by public and private companies - Field of legal, medical, agriculture etc. Making knowledge base common standard defining general knowledge and structure - Use in artificial intelligence speakers, counseling chatbots, etc. Establishing guidelines for knowledge base to be able to link and utilize data Provide technical support to ministries, agencies, and companies (2018 )

Reference] Technical standard trend of AI field (International standard) ISO / IEC JTC1 General Assembly passed the establishment of SC 42 (Subcommittee 42), a new AI subcommittee proposed by the United States (October, 2017) * JTC 1 (Joint Technical Committee 1) is a standards committee in the field of information and communication (ICT) (base standard) terminology, knowledge standard, reference structure, framework, ontology (calculation method) Fuzzy logic, neural network, machine learning (Reliability) Consider the balance of security, privacy, harm and benefits (National Standard) In response to the establishment of ISO / IEC JTC 1 / SC 42, to be configured * In addition, a separate AI PG(Project Group) will be established in TTA supported by NIA

V. Social change by data 26

1.1 Social change by data Indust ry Technolo gy Unstructured data analysis Medical SNS Voice Video Document Societ y Big data Order made health care - Public health alert service by National Health agency AI technology graft Traffic Big data Improve traffic flow - Optimize midnight bus route Data collect Combination Machine Result uses Learning Open source platformManufacture Big data Improve work conditions and reduce costs - Smart factory CPS(Cyber Physical System)

Thank you 28