I'm Seong-ho Shin, working for KST in South Korea. I'm in charge of developing DataON, which is the Korea research data platform. So today I'm going to talk about the DataON. Before starting, I want to show two pictures introducing myself, because this is not a live talk.

This is me. I am now in my office, and the other picture here shows my office space. Most workers in my organization are working together in a workspace divided into each department.

Okay, now let me get into my talk. My organization, KST, is a Korean National Institute for Researching Science and Technology Information.

KST is a leading institute for Open Science of Korea. It is researching and also building various kinds of platforms and web services related to information, data analysis, and so on. Its aim is to help Korea to support building open science environment.

KST has DataON for research data sharing, ScienceON for patent publications, reports, and AccessON for open access articles.

We also have data analysis division here and supercomputing infrastructure. Researchers can use computer power for their researches from KST's computer.

This is a brief introduction on KST. The main topic of my presentation is about our research data platform. So, I need to first mention about Korea government strategy for Korea research data sharing because the strategy has built the platform.

Korea government has established the strategy for research data 2017. The strategy consists of four key elements such as legal system, human resources, infrastructure, and research community.

The goal is promoting sharing and utilization of national research data through knowledge assetization. DataON is a way of knowledge assetization and also the realization of infrastructure element.

DataON is linking with domain specific research data platform centers and their data platforms step by step. So, in the future, most research data of Korea will be loaded in DataON and researchers can search what they need through DataON.

There are various kinds of research data like bio data, genetic data, LHC data, telescope data, and so on.

But considering information and technology domain, I want to say that the boundary of research data is not that clear. Because recently, information and technology researchers work into most academic fields.

The opponent situation also happens. So, we define that research data is essential and objective factual data for reproduction of research results obtained through various experiments, observation, surveys, and analysis of national R&D project.

We also consider data management plan called, in short, DMP. The goal of DMP is to consider the many aspects of data management, metadata generation, data preservation, and analysis before the project begins.

This ensures that data are well managed in the present and prepared for preservation in the future. From now, I will introduce DataON, especially for its services and functions.

DataON is a research data management and sharing platform. CAST is not building much research data, but just a few. Then, where is research data? Research data is stored in researchers' servers or PCs.

From this, one of the main functions of DataON is data gathering or linkage. Research data stored in DataON needs to be easy to search and easy to use.

So, DataON should be a portal service and have several services for usage encouragement. This is the conceptual diagram of DataON. These are service data preservation, convenient search, analysis environment, and online community and so on.

It has also DMP, data repository, data registration, and also data linkage.

In order to collect research data, DataON distributes an institutional data repository software to research organization having research data. We also have data connection with global research platforms like OpenAIRE EU, ARTC Australia, and ARCOS Japan.

From this slide, some features of DataON will be mentioned showing its web pages. First one is data registration functions, basic information, information on who visitors, file information, copyright, and extra information.

Also, this is the data search service. DataON provides basic search, advanced search, facet search, map search, and search within users.

This is the most attractive service of DataON. There are two ways of data analysis. One is workflow. This is an easy way to use analysis environment.

The other is the command line interface based on JupyterLab. This slide shows the first one, workflow. Suppose, as a researcher, I have data set but not good at coding to analyze the data set.

But there are so many open source code libraries of software on the internet. I can find what libraries I need for task and register them into DataON.

And by drag and drop and some compilation of input output parameters, I can activate the software and analyze data set on DataON. Very easy to use. This is the JupyterLab embedded in DataON for who are familiar with coding.

It supports distributed deep learning, machine learning framework, and also Hadoop ecosystem. 16 cores, CPU, one Nvidia V100 GPU, and one terabyte storage space are basically provided per users.

OK, finally, we reached to the end. I've mentioned so far that DataON designed to support and contribute Korea Open Science.

Based on Korea government strategy, DataON has been built from 2018 and opened its formal service in January 2020. It preserves researchers' data sets and supports them sharing data sets with each other by using functions and services of DataON.

This is all I prepared, sorry for not live talk, but I tried to do my best to answer your question via my email. Thank you for listening.