Keungoui Kim
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Keungoui Kim has a total of 54 co-authors. Here are the top co-authors by number of shared publications:
- Dieter Franz Kogler
University College Dublin
14 shared publications - Junseok Hwang
Seoul National University
8 shared publications - Junmin Lee
Pusan National University
7 shared publications - Changjun Lee
School of Convergence, Sungkyunkwan University, Seoul 03063, Republic of Korea
5 shared publications - Jisoo Hur
TUM School of Management Technical University of Munich, Munich, Germany
5 shared publications - Hyunha Shin
National Institute of Green Technology , Seoul , Republic of Korea
4 shared publications
Latest Publications
- Open Access
- 19 February 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (2) , e0341005
Knowledge flows from science to AI technology: Identifying core and brokerage technological roles
The rapid advancement of artificial intelligence (AI) technologies has not only driven convergence with diverse technological domains but also swiftly spread across various industrial sectors. As a knowledge-intensive field, AI is particularly shaped by the flow of knowledge from scientific research to technological development, yet remains insufficiently examined in a systematic and structural way. This study addresses this gap by investigating science-to-technology knowledge flow that underpins AI’s technological evolution. We propose a semantic science-technology exploration framework specifically designed for the AI domain, consisting of the two stages: technology classification and semantic topic exploration. First, AI patents are classified into four categories using centrality measures derived from a CPC co-occurrence network. Then, we extract abstracts from both patents and their cited scientific publications to apply BERTopic modelling and generate topic labels using generative AI. Analyzing AI-related patents filed from 2002 to 2021, we trace key technological trends and elucidate the structural pathways of knowledge flow science to technology. The findings offer practical implications for corporate R&D strategies and innovation policy design in the era of AI.
- 19 February 2026
- Published by Public Library of Science (PLoS) in PLOS ONE
- Vol. 21 (2) , e0341005
Open AccessKnowledge flows from science to AI technology: Identifying core and brokerage technological roles
The rapid advancement of artificial intelligence (AI) technologies has not only driven convergence with diverse technological domains but also swiftly spread across various industrial sectors. As a knowledge-intensive field, AI is particularly shaped by the flow of knowledge from scientific research to technological development, yet remains insufficiently examined in a systematic and structural way. This study addresses this gap by investigating science-to-technology knowledge flow that underpins AI’s technological evolution. We propose a semantic science-technology exploration framework specifically designed for the AI domain, consisting of the two stages: technology classification and semantic topic exploration. First, AI patents are classified into four categories using centrality measures derived from a CPC co-occurrence network. Then, we extract abstracts from both patents and their cited scientific publications to apply BERTopic modelling and generate topic labels using generative AI. Analyzing AI-related patents filed from 2002 to 2021, we trace key technological trends and elucidate the structural pathways of knowledge flow science to technology. The findings offer practical implications for corporate R&D strategies and innovation policy design in the era of AI.
- 4 February 2026
- Published by Springer Nature in American Journal of Criminal Justice
- Vol. 51 (5) , 1247-1275
Understanding the Spatial Co-Occurrence of Different Crime Types and Brokerage Crime in NYC Block Groups
- 4 February 2026
- Published by Springer Nature in American Journal of Criminal Justice
- Vol. 51 (5) , 1247-1275
Understanding the Spatial Co-Occurrence of Different Crime Types and Brokerage Crime in NYC Block Groups
- 29 September 2025
- Published by Springer Nature in Scientometrics
- Vol. 130 (10) , 5793-5820
Mapping regional knowledge spaces via BERTopic and BERT similarity: the semantic proximity of quantum science in the overall scientific publication landscape
- 29 September 2025
- Published by Springer Nature in Scientometrics
- Vol. 130 (10) , 5793-5820
Mapping regional knowledge spaces via BERTopic and BERT similarity: the semantic proximity of quantum science in the overall scientific publication landscape
- Open Access
- 23 September 2025
- Published by Walter de Gruyter GmbH in ZFW – Advances in Economic Geography
- Vol. 69 (2-3) , 107-125
Science space: the evolution of scientific knowledge specialisations across European regions
This study extends evolutionary economic geography to science by mapping how regional scientific capabilities emerge and evolve across Europe. Using Web of Science publications (2000–2017) geo-coded to 1,216 regions in 35 countries and classified into 228 subjects, we construct a pan-European “Science Space” based on subject co-occurrence and relatedness, and test whether relatedness density, i.e., the embeddedness of a subject in a region’s existing portfolio, predicts subsequent entry (Revealed Scientific Advantage ≥ 1). Network evidence shows Europe’s science system becoming more interdisciplinary, with technology-adjacent subjects (e.g., nanoscience, robotics, computer science) gaining centrality, while Life Sciences & Biomedicine remain dominant by volume. Econometric results (pooled OLS and GLM for binary entry, with region/subject/period fixed effects and controls for the regional economy and knowledge base) indicate that higher relatedness density significantly raises the likelihood of scientific entry. The effect is stronger in non-metropolitan regions and when a subject’s initial RSA is very low, consistent with relatedness seeding new capabilities rather than merely consolidating near-threshold strengths. These findings generalise the principle of relatedness from technology to science and advise regional innovation policy to prioritise adjacent scientific opportunities, invest in bridging infrastructures, and design interdisciplinary platforms where relatedness density is high but specialisation has not yet emerged.
- 23 September 2025
- Published by Walter de Gruyter GmbH in ZFW – Advances in Economic Geography
- Vol. 69 (2-3) , 107-125
Open AccessScience space: the evolution of scientific knowledge specialisations across European regions
This study extends evolutionary economic geography to science by mapping how regional scientific capabilities emerge and evolve across Europe. Using Web of Science publications (2000–2017) geo-coded to 1,216 regions in 35 countries and classified into 228 subjects, we construct a pan-European “Science Space” based on subject co-occurrence and relatedness, and test whether relatedness density, i.e., the embeddedness of a subject in a region’s existing portfolio, predicts subsequent entry (Revealed Scientific Advantage ≥ 1). Network evidence shows Europe’s science system becoming more interdisciplinary, with technology-adjacent subjects (e.g., nanoscience, robotics, computer science) gaining centrality, while Life Sciences & Biomedicine remain dominant by volume. Econometric results (pooled OLS and GLM for binary entry, with region/subject/period fixed effects and controls for the regional economy and knowledge base) indicate that higher relatedness density significantly raises the likelihood of scientific entry. The effect is stronger in non-metropolitan regions and when a subject’s initial RSA is very low, consistent with relatedness seeding new capabilities rather than merely consolidating near-threshold strengths. These findings generalise the principle of relatedness from technology to science and advise regional innovation policy to prioritise adjacent scientific opportunities, invest in bridging infrastructures, and design interdisciplinary platforms where relatedness density is high but specialisation has not yet emerged.
- 1 September 2025
- Published by Springer Nature in Evolutionary and Institutional Economics Review
- Vol. 22 (2) , 485-509
The firm-location nexus: examining inventive footprint and technological evolution of Nortel over the course of its “boom and bust”
- 1 September 2025
- Published by Springer Nature in Evolutionary and Institutional Economics Review
- Vol. 22 (2) , 485-509
The firm-location nexus: examining inventive footprint and technological evolution of Nortel over the course of its “boom and bust”
- 19 August 2025
- Published by Taylor & Francis in Innovation
- Vol. 28 (4) , 1-19
Unravelling the impact of design-engineering capability on firm performance
- 19 August 2025
- Published by Taylor & Francis in Innovation
- Vol. 28 (4) , 1-19
Unravelling the impact of design-engineering capability on firm performance
- 30 June 2025
- Published by Korean Institute of Information Technology in The Journal of Korean Institute of Information Technology
- Vol. 23 (6) , 31-40
A Study on Data Preprocessing and Prompt Engineering Methodologies Specialized for the Interpretation of Automobile Law Documents
자동차 설계 및 개발 과정에서 개발자는 법과 규제를 준수하기 위해 담당 법률 전문가와 지속적으로 소통하고 협력해야 한다. 그러나 기업 내 법률 전문가의 수가 제한적이라는 현실에서는 다수의 개발자가 적시에 법률 자문과 검토를 받기 어렵다. 본 연구는 법률 비전문가인 자동차 개발자가 생성형 AI를 활용하여 법률 및 규제를 검토하고 해석할 수 있도록, 자동차 법률 문서에 특화된 데이터 전처리 및 프롬프트 엔지니어링 방법론을 제안한다. 제안된 방법론은 법률 문서의 구조적 정보를 활용하여 검색 효율성을 높이는 Parent-Child Chunking 기법과 답변 정답률을 향상시키기 위한 프롬프트 엔지니어링 기법을 포함한다. 이러한 접근 방식으로 구축된 RAG 서비스는 자동차 개발자가 법규 관련 업무를 보다 효율적이고 자율적으로 수행하는데 기여할 수 있을 것으로 기대된다.
- 30 June 2025
- Published by Korean Institute of Information Technology in The Journal of Korean Institute of Information Technology
- Vol. 23 (6) , 31-40
A Study on Data Preprocessing and Prompt Engineering Methodologies Specialized for the Interpretation of Automobile Law Documents
자동차 설계 및 개발 과정에서 개발자는 법과 규제를 준수하기 위해 담당 법률 전문가와 지속적으로 소통하고 협력해야 한다. 그러나 기업 내 법률 전문가의 수가 제한적이라는 현실에서는 다수의 개발자가 적시에 법률 자문과 검토를 받기 어렵다. 본 연구는 법률 비전문가인 자동차 개발자가 생성형 AI를 활용하여 법률 및 규제를 검토하고 해석할 수 있도록, 자동차 법률 문서에 특화된 데이터 전처리 및 프롬프트 엔지니어링 방법론을 제안한다. 제안된 방법론은 법률 문서의 구조적 정보를 활용하여 검색 효율성을 높이는 Parent-Child Chunking 기법과 답변 정답률을 향상시키기 위한 프롬프트 엔지니어링 기법을 포함한다. 이러한 접근 방식으로 구축된 RAG 서비스는 자동차 개발자가 법규 관련 업무를 보다 효율적이고 자율적으로 수행하는데 기여할 수 있을 것으로 기대된다.
- 30 November 2024
- Published by Korean Institute of Information Technology in The Journal of Korean Institute of Information Technology
- Vol. 22 (11) , 167-180
Reseach on Prompmt Engineering for Generative AI to Create Business Data Analysis Reports
본 논문에서는 퍼포먼스 마케팅을 중심으로 비즈니스 데이터 분석 리포팅을 최적화하기 위한 생성형 AI 기반 프롬프트 구조를 제안하고자 한다. 제안된 방식에는 퍼포먼스 마케팅 주요 성과 지표로 구성된 일일 리포팅 데이터셋을 전처리 과정과 테이블 형식의 데이터를 자연어 텍스트로 변환하는 'Table-to-Text' 방식이 포함되었다. 해당 도메인에서 주로 활용되는 질의에 대한 적정한 답변을 생성할 수 있도록 다양한 프롬프트 기법을 조합하고 검토하였으며 ROUGE Score와 자체 체크리스트를 활용한 평가 결과를 토대로 최적의 프롬프트를 선정하였다. 본 논문은 퍼포먼스 마케팅 분야에서의 프롬프트 엔지니어링을 연구한 점에서 기존 논문들이 다루지 않았던 분야를 탐구했다. 특히, 테이블 데이터 해석과 데이터의 주요 포인트를 반영한 코멘트를 동시에 생성하는 프롬프트 구조를 제시함으로써, 기존 연구와 차별화되는 접근 방식을 제안한다. 본 논문에서 제안한 프롬프트를 통해 퍼포먼스 리포트 작성시의 객관성을 높이고 시간 및 비용을 절약시킬 수 있을 것으로 기대된다
- 30 November 2024
- Published by Korean Institute of Information Technology in The Journal of Korean Institute of Information Technology
- Vol. 22 (11) , 167-180
Reseach on Prompmt Engineering for Generative AI to Create Business Data Analysis Reports
본 논문에서는 퍼포먼스 마케팅을 중심으로 비즈니스 데이터 분석 리포팅을 최적화하기 위한 생성형 AI 기반 프롬프트 구조를 제안하고자 한다. 제안된 방식에는 퍼포먼스 마케팅 주요 성과 지표로 구성된 일일 리포팅 데이터셋을 전처리 과정과 테이블 형식의 데이터를 자연어 텍스트로 변환하는 'Table-to-Text' 방식이 포함되었다. 해당 도메인에서 주로 활용되는 질의에 대한 적정한 답변을 생성할 수 있도록 다양한 프롬프트 기법을 조합하고 검토하였으며 ROUGE Score와 자체 체크리스트를 활용한 평가 결과를 토대로 최적의 프롬프트를 선정하였다. 본 논문은 퍼포먼스 마케팅 분야에서의 프롬프트 엔지니어링을 연구한 점에서 기존 논문들이 다루지 않았던 분야를 탐구했다. 특히, 테이블 데이터 해석과 데이터의 주요 포인트를 반영한 코멘트를 동시에 생성하는 프롬프트 구조를 제시함으로써, 기존 연구와 차별화되는 접근 방식을 제안한다. 본 논문에서 제안한 프롬프트를 통해 퍼포먼스 리포트 작성시의 객관성을 높이고 시간 및 비용을 절약시킬 수 있을 것으로 기대된다
- 4 October 2024
- Published by Taylor & Francis in Justice Quarterly
- Vol. 42 (7) , 1300-1320
A Machine Learning Approach to Analyzing Crime Concentration: The Case of New York City
- 4 October 2024
- Published by Taylor & Francis in Justice Quarterly
- Vol. 42 (7) , 1300-1320
A Machine Learning Approach to Analyzing Crime Concentration: The Case of New York City
- 29 July 2024
- Published by Oxford University Press (OUP) in Science and Public Policy
- Vol. 51 (6) , 1257-1266
Exploring the interplay between social distancing, innovation adoption, and privacy concerns amid the COVID-19 crisis
Using the case of coronavirus disease-2019 (COVID-19) as a natural experiment, this study discusses how innovation policy should be designed in times of crisis when public interest and privacy concerns collide. For this purpose, the causal relationship between social distancing enforcement and innovation adoption in consideration of privacy concerns is explored by using national-level survey data. Our key findings from econometric analysis are as follows: first, the implementation of social distancing policies leads to an increase in the adoption of wearable devices. Second, the strengthening of social distancing has caused a decrease in individual privacy concerns. Finally, a decrease in individual privacy concerns leads to an increase in wearable device adoption. Social distancing during the COVID-19 not only accelerated innovation adoption but also led a decrease in privacy concerns. Our results suggest that government should play a role in safe guarding public privacy when individuals may let their guard down during times of crises.
- 29 July 2024
- Published by Oxford University Press (OUP) in Science and Public Policy
- Vol. 51 (6) , 1257-1266
Exploring the interplay between social distancing, innovation adoption, and privacy concerns amid the COVID-19 crisis
Using the case of coronavirus disease-2019 (COVID-19) as a natural experiment, this study discusses how innovation policy should be designed in times of crisis when public interest and privacy concerns collide. For this purpose, the causal relationship between social distancing enforcement and innovation adoption in consideration of privacy concerns is explored by using national-level survey data. Our key findings from econometric analysis are as follows: first, the implementation of social distancing policies leads to an increase in the adoption of wearable devices. Second, the strengthening of social distancing has caused a decrease in individual privacy concerns. Finally, a decrease in individual privacy concerns leads to an increase in wearable device adoption. Social distancing during the COVID-19 not only accelerated innovation adoption but also led a decrease in privacy concerns. Our results suggest that government should play a role in safe guarding public privacy when individuals may let their guard down during times of crises.