content area

Leading a data-driven culture via creation
of data ecosystems

UNIST Start-up
Core.Today Co., Ltd

The biggest change in the 4th Industrial Revolution is convergence without the boundaries of super connection and super intelligence. Its basis is data. The competence to handle the data with AI is considered a necessary capability in the super connected society. There is a startup that has presented a grand vision that catches up with such flow and establishes a virtuous circle ecosystem to utilize data and make a new data culture. Core.Today Co., Ltd is recognized as the first professional company engaged in AI based data processing in Korea.
  • Story by. Editorial Room   Photo by. Kim Beon-Ki

“We do not stop with a simple understanding of AI through theories but try to make a ‘Data Story Telling Platform’ that provides the space, tools, and sharing environment for anyone to analyze original data or processed data. I hope that this platform can grow into a data ecosystem for the function of contact point and the mutual virtuous circle of a data creator who can explore the reliable data based story naturally and analyze the subscribers to understand the information and the data.”

“We do not stop with a simple understanding of AI through theories but try to make a ‘Data Story Telling Platform’ that provides the space, tools, and sharing environment for anyone to analyze original data or processed data. I hope that this platform can grow into a data ecosystem for the function of contact point and the mutual virtuous circle of a data creator who can explore the reliable data based story naturally and analyze the subscribers to understand the information and the data.”

Developed AI based data model in diverse fields

In the analogue era, tangible assets of land, labor, and capital were important, but in the digital era, intangible assets such as data or algorithm occupy their position. The core competitiveness today is how much we can understand and handle such cyber assets well. Its basis is AI. Our mind is moved by this statement of future scientists, ‘The annual salary will be determined by how well we can cooperate with AI.’ Yet, many cases show that people simply learned the theory on AI and stopped. They find it difficult to set a direction where and how they can utilize the theory in reality. Core.Today Co., Ltd considers that they would play the role of a guide for that reality. Core.Today is a professional company that processes and analyzes original data to provide various data models that conform to necessary purposes. Professor Bongsoo Jang at the Department of Mathematical Sciences UNIST established the company in 2016 as a teacher startup. Currently, post-PhD researcher Kim Kyung-Hoon presides as a representative who worked with Professor Jang. In its early period of startup, they focused mainly on the technical development of natural language processing that summarizes Korean news. The natural language processing algorithm handles the mutual reaction between the computer and the human language, which is one of core functions of AI. To extract and utilize meaningful information from massive texts, technical researches into deep learning and deep learning-based natural language processing have been carried out actively. Core.Today has continued many R&Ds into algorithm from the early period. The accumulated technology has become a strong foundation to support today’s Core.Today.
Regardless of the short history of accomplishment, the path Core.Today has walked is exceptional. Core.Today drew attention through the establishment of a service ‘LAWBOT,’ which is an optimized search engine developed in 2018 for judicial precedents. LAWBOT is a search engine to extract, process and analyze the data for judicial precedents among public data. It received a fervent response as it was able to search laws around the meaning not simply the word and to identify the relationship between judicial precedents through the visual engine. Based on the solid technology, in 2019 Core.Today was selected as an enterprise to supply data vouchers under the supervision of the Ministry of Science and ICT. Last year, Core.Today participated in the contest named First Great Star Solver Platform’ that is suggested by giant enterprises and solved by startups and was selected as a solver at the final with the ‘Method to Predict Viewer’s Ratings’ presented by KBS. At this time, Core.Today deduced a genre through a synopsis and analyzed the character for each casting to propose an algorithm to predict the viewer’s ratings for each candidate actor in a drama. Core.Today received recognition for the fact that AI learning including internal information in relation to the production of a drama will raise the accuracy of prediction for the performance of the drama and help determining actors in the preliminary planning stage.

Toward a data ecosystem of virtuous circle created by subscriber and creator

The effort to foster talented personnel is the most outstanding step Core.Today has taken. The CEO Kim Kyung-Hoon has especially concentrated on fostering data creators who can produce data stories. He said, “While there exist large volumes of mixed data, we need the contents to transfer the message clearly. A person who makes such contents is the data creator. As the creator of Youtube, the data creator utilizes individual technology, knowledge and ideas in diverse ways to make the contents that subscribers can enjoy. We name these contents as data story.” Currently, Core.Today has been operating a training course named ‘AI Learning School for Data Creators’ through a MOU with Ulsan ICT Promotion Agency. The education proceeds with practical works to experience diverse methods of data utilization ranging from Python Basic Education to make data into a story to the AI intensive course. The employee Kim Do-Hee who has been working as a data creator in Core.Today and Ahn Kuk-Moon as a data engineer are also the talented personnel who graduated in the courses on AI and data through the learning school.
The CEO Kim said, “Utilizing the data of text, image and time series to develop a model in diverse areas and foster data creators is the process for the making of an example for the utilization of AI and laying the stepping stone toward the goal to share the algorithm with all.” Namely, its goal is to develop such data processing technique into API (Application Programming Interface) and to make a python library for easy open application of the API. Some software has already been disclosed as open source. The data tool of Core. Finance is a representative case of handling financial data. Moreover, Core. Today is scheduled to disclose the function in relation to text, image and time series that are currently in development or in future development for many data creators to utilize conveniently through ‘Core. Today Open Source Storage (https://github.com/coredottoday),’ which will be launched officially in the second half of this year. In addition, the CEO Kim revealed the vision of Core.Today, “We plan to make a ‘Data Story Telling Platform’ that will not stop at just learning the theory of AI but provide the space, tools and sharing environment to analyze original data or processed data. I hope that this platform will grow into a data ecosystem and serve the function of a contact point with data creators and provide a mutual virtual circle to analyze both the subscriber who searches the reliable data-based story naturally and learns the information and the data that goes into creating the story.”
‘Core.Today suggests the vision, ‘Telling a story of message for data.’ and takes the lead to create a new data culture. The day is not too far when the AI-based data story telling created by Core. Today will form a vast forest .”

MINI INTERVIEW
  • Space to work with initiative and grow on one’s own Kim Do-Hee, Data Creator; Leave of absence in the senior year of the School of Computer Science and Engineering at UNIST

    “After the military service, I thought I should experience various works and find my own path while I looked at my friends who knew their goals and were moving forward. Then, I took a leave of absence for 2 months during the last winter vacation to create some momentum by participating in ‘Development of AI Model Related to Box-Office Hit of KBS Dramas’ through the education program (CUop) in connection with the industry-academy cooperation and began working with Core.Today. It was fun and joyful to collect the data in the field of sports and games and make diverse data stories. Now, I am establishing a nerve network model for the collection and analysis of data to predict the box-office hit of dramas. It is not intended to simply predict the box-office hit of dramas but to convert the natural language process to API and disclose it as an open source that many data creators can utilize. I consider every aspect from the planning stage to its realization and work with initiative. I am satisfied as I can grow to a better data creator.”

  • This path was found after going around for 7 years; I am having fun working on it even if it is not my major. Ahn Kook-Moon, Data Engineer, Graduated from the Department of Eco-friendly Energy Engineering at UNIST

    “Before I came to work here, I had worked to set up the goals and make the indices to evaluate and manage performances for almost 7 years. However, I found it less fun to repeat work every year rather than attempting a new method. Then, I happened to meet our CEO and began working after the completion of 160 hours in the leaning school. Later, I learned several practical works of AI development. Now, I am developing ‘AI Painter’ through the R&D project for new technology, which is supported by the City of Ulsan. The project is to maintain the basic shape of original images and utilize the AI algorithm to make a new image by applying the desired style of painting. It has the goal to provide the public with easy access to AI and art. We plan to hold a related exhibition in the second half of this year. Even the present work is different from the major I pursued in my college years, but I really work with pleasure. I think that it is meaningful to create valuable stories with data, which will grow more.”