빅데이터와 한국 고대사 연구경향

서호준 ( Seo¸ Ho-joon ) 대구사학회 2021 대구사학 Vol.144 No.-

2016년 다보스 포럼에서 4차 산업혁명이 화두로 등장한 이래 다양한 학문영역에서 4차 산업혁명에 대한 대응을 서두르고 있다. 역사학을 포함한 인문학 분야에서도 4차 산업혁명에 대한 관심이 고조되고 있지만, 아직까지 이에 대한 대응은 지지부진한 상황을 면치 못하고 있는 것은 생각된다. 본 연구는 이러한 4차 산업혁명과 역사학간의 연결고리를 찾고자 하는 의도에서 비롯되었다고 할 수 있으며, 보다 구체적으로는 4차 산업혁명의 핵심 기술로 주목받고 있는 빅데이터 기법을 어떻게 하면 역사학, 특히 한국 고대사 분야에 접목시킬 수 있을 것인가하는 것이 바로 연구의 목적이라할 수 있다. 사실 역사학은 방대한 사료를 연구대상으로 삼고 있다는 점에서 가장 오랜 데이터과학으로서의 성격을 갖고 있다. 그러나 상관성을 추구하는 빅데이터와 인과성에 대한 탐문을 목적으로 하는 역사학간에는 그 접점을 찾기가 쉽지 않을 것처럼 보이는 것도 사실이다. 그렇지만 이러한 시각은 빅데이터의 위상을 지나치게 과대평가한 것으로 역사연구를 수행하는 하나의 도구로서만 빅데이터를 활용한다면 역사연구는 한층 효율적으로 이루어질 수 있을 것으로 생각되었다. 본 연구는 이와 같은 빅데이터와 한국 고대사 연구간의 접목을 위해 해방 이후 2010년대까지 약75년에 이르는 한국 고대사의 연구경향을 크게 7개 시기로 구분하여 텍스트 마이닝과 소셜 네트워크 분석을 시도하였다. 빅데이터 기반 분석결과를 한국 고대사 연구경향에 관한 기존의 논의결과와 비교고찰하였다. 본 연구의 한국 고대사 연구경향 분석에서 나타난 결과는 다음의 몇 가지로 정리될 수 있다. 첫째, 논저의 수가 크게 증가하고 있는 최근으로 올수록 그 정확성이 높아지는 현상이 관찰된다는 것이다. 둘째, 특정 국가사의 활성화 정도를 직관적으로 논의하는 경우 텍스트 네트워크 분석을 보충적 장치로 활용할 수 있다는 점이다. 셋째, 텍스트 네트워크 분석은 기본적으로 키워드(노드) 분석을 위한 것이므로 문장단위 분석이나 고차원적인 개념분석에는 약점을 갖고 있는 것으로 나타난 점이다. 사실 사료에 대한 면밀하고 신중한 검토가 요망되는 고대사 연구를 계량적으로만 분석한다면 오히려 역사적 진실을 왜곡하게 될 위험성마저 내포하고 있는 것으로 볼 수 있다. 그러나 빅데이터 기반 연구는 이와 같은 단순 계량적 접근을 넘어 사료나 연구경향 논문들을 대상으로 한 키워드 도출을 통해 가상의 네트워크를 구성하여 네트워크내 키워드들의 위상과 영향력 그들 상호간의 관련성을 파악하고자 하였다. 실증과정에서 적지 않은 한계를 노출하긴 하였지만, 본 연구가 제시한 분석의 프레임 자체는 양질의 후속연구를 태동시키기 위한 연결고리로서는 충분히 기능할 수 있을 것으로 기대된다. Since the 4th Industrial Revolution appeared as a hot topic at the Davos Forum in 2016, various academic fields have been in a hurry to respond to the 4th Industrial Revolution. Interest in the Fourth Industrial Revolution is also increasing in the humanities field including history, but it is thought that the response to this is still sluggish. The purpose of this study is to find a link between the 4th Industrial Revolution and history, and more specifically, how to incorporate big data techniques into the field of history, especially ancient Korean history In fact, historical science has the longest character of data science in that it uses vast historical sources for research. However, it is also true that it will not be easy to find the point of contact between big data seeking correlation and history studies aiming to investigate causality. However, this view overestimated the status of big data, and it was thought that if big data was used only as a tool for conducting historical research, historical research could be carried out more efficiently. This study performed big data-based text network analysis to combine such big data with research on ancient Korean history. The research trend of Korean ancient history, which spans about 75 years from independence to the 2010s, is largely divided into 7 periods, and attemped text mining and social network analysis. The results of analysis based on big data were compared with the results of previous discussions on the research trend of ancient Korean history. In fact, if the research on ancient history, which requires close and careful consideration of historical sources, is analyzed only quantitatively, it can be seen as a risk of distorting the historical truth. However, the big data-based research went beyond such a simple quantitative approach to construct a virtual network by deriving keywords for historical sources or research trend papers to understand the status and influence of keywords in the network and their relationship to each other. Although some limitations have been exposed during the demonstration process, the frame itself of the analysis presented by this study is expected to function fully as a link to initiate quality follow-up studies of good quality. (Korea Credit Guarantee Fund / horays@naver.com)

INFLUENCES OF CULTURAL AND SOCIAL RELATIONSHIP VARIABLES ON ELECTRONIC WORD-OF-MOUTH BEHAVIOR

Hojoon Choi,Kyung Yul Lee,Temple Northup 글로벌지식마케팅경영학회 2020 Global Marketing Conference Vol.2020 No.11

Query-based zero-shot detection

HoJoon Song(송호준),Ganghun Lee(이강훈),Byoung-Tak Zhang(장병탁) 한국정보과학회 2020 한국정보과학회 학술발표논문집 Vol.2020 No.7

Low Voltage Ride Through(LVRT) Control Strategy of Grid-connected Variable Speed Wind Turbine Generator System

Hojoon Shin,Hyun-Sam Jung,Seung-Ki Sul 전력전자학회 2011 ICPE(ISPE)논문집 Vol.2011 No.5

In this paper, a Low Voltage Ride Through (LVRT) control strategy for variable speed full scale Wind Turbine Generator System (WTGS) is proposed. Grid-connected wind power system should satisfy LVRT requirement when grid fault occur, and the variable speed full scale WTGS has advantage of satisfying LVRT compared with the fixed speed WTGS. The proposed LVRT control strategy satisfies not only LVRT requirement but also stabilizing a DC link voltage regardless of reduced DC link capacitance. During the grid faults, especially three phase to ground fault, the grid side converter cannot control the DC link voltage because it is impossible to control grid side power. In this situation, the DC link voltage can be controlled by the generator side inverter using the proposed DC link voltage control scheme. The proposed control scheme can keep the power balance between grid and generator side power in any case of grid fault condition for LVRT. The compliance to Grid Codes has been evaluated by computer simulations. Though the experimental results, it is verified that the proposed method has satisfied the LVRT requirements and DC link voltage control specification simultaneously.

Performance estimation of freeze protection system for outdoor fire piping by using AI algorithm

Hojoon Cho,Sangmin Seo,Chinseok Heo,Junjae Kwak,Yongbae Kim,Jinsoo Park,Sangjun Lee,임성식 대한기계학회 2023 JOURNAL OF MECHANICAL SCIENCE AND TECHNOLOGY Vol.37 No.10

In the present study, the performance of the metal heater-based freeze prevention system was predicted with respect to the major design variables of outdoor fire piping, including the fire pipe diameter, outer temperature, and insulation thickness. To this end, CFD analysis was conducted to obtain water center temperature data along the pipe-length direction with respect to the major design variables. Subsequently, four AI algorithms, including the deep neural network, decision tree, random forest, and support vector machine, were trained with the collected data, and their prediction performance was compared. Further, each algorithm, once trained, was tested for its ability to make reasonable predictions for the conditions that it had not been trained with. Overall, the deep neural network model exhibited the best prediction performance for both interpolation and extrapolation data. As a result, the model was determined to be the most suitable for the prediction of the water temperature.

Integrity Monitoring for Drone Landing in Urban Area using Single Frequency Based RRAIM

Hojoon Jeong,Bu-Gyeom Kim,기창돈 사단법인 항법시스템학회 2022 Journal of Positioning, Navigation, and Timing Vol.11 No.4

In this paper, we developed a single frequency-based RRAIM to monitor integrity of the UAM landing vertically in urban area with only low-cost single-frequency GPS receiver. Conventional dual-frequency RRAIM eliminates ionospheric delay through a combination of frequencies. In this study, ionospheric delay was directly modeled. Drift error of residual ionospheric delay is modeled using the previously studied result on ionospheric rates of change. To verify the performance of the proposed RRAIM algorithm, a simulation of vertical landing UAM in urban area was conducted. It was assumed that the protection level at the initial position was calculated through SBAS correction data. During vertical landing, integrity monitored by receiver alone without external correction data. In the 60 sec simulation, the protection level of the proposed RRAIM compared to the conventional RRAIM was calculated to be 140% due to the accumulated ionospheric delay error. Nevertheless, it was confirmed that the final vertical protection level meeting the requirements of LPV-200, which cannot be achieved with single frequency GPS receiver alone.

New records of five taxa of unarmored and thin-walled dinoflagellates from brackish and coastal waters of Korea

Hojoon Choi,Minji Cho,Sunju Kim 한국환경생물학회 2021 환경생물 : 환경생물학회지 Vol.39 No.4

Unarmored and thin-walled dinoflagellates were collected from brackish and coastal waters of Korea from August 2019 to August 2021. A total of 10 species belonging to orders Sussiales and Gymnodiniales were isolated and established as clonal cultures. Of them, five species (Biecheleria brevisulcata, Lepidodinium chlorophorum, Karlodinium decipiens, Kirithra asteri, and Wangodinium sinense) are newly recorded in Korea and examined using a light microscope (LM) and a scanning electron microscope (SEM). Their molecular phylogeny was inferred from LSU rDNA sequences. Here, we present taxonomic information, morphological features, and molecular phylogenetic positions of these unrecorded dinoflagellate species.

Design of Orbit Simulation Tool for Lunar Navigation Satellite System

Hojoon Jeong,Jaeuk Park,Junwon Song,Minjae Kang,Changdon Kee 사단법인 항법시스템학회 2023 Journal of Positioning, Navigation, and Timing Vol.12 No.4

Lunar Navigation Satellite System refers to a constellation of satellite providing PNT services on the moon. LNSS consists of main satellite and navigation satellites. Navigation satellites orbiting around the moon and a main satellite moves the area between the moon and the L2 point. The navigation satellite performs the same role as the Earth’s GNSS satellite, and the main satellite communicates with the Earth for time synchronization. Due to the effect of the non-uniform shape of the moon, it is necessary to focus on the influence of the lunar gravitational field when designing the orbit simulation for navigation satellite. Since the main satellite is farther away from the moon than the navigation satellite, both the earth’s gravity and the moon’s gravity must be considered simultaneously when designing the orbit simulation for main satellite. Therefore, the main satellite orbit simulation must be designed through the three-body problem between the Earth, the moon, and the main satellite. In this paper, the orbit simulation tool for main satellite and navigation satellite required for LNSS was designed. The orbit simulation considers the environment characteristics of the moon. As a result of comparing long-term data (180 days) with the commercial program GMAT, it was confirmed that there was an error of about 1 m.

Wake Characteristics of Vane-Type Vortex Generators in a Flat Plate Laminar Boundary Layer

HoJoon Shim,Young-Hee Jo,Kyoungsik Chang,Ki-Jung Kwon,Seung-O Park 한국항공우주학회 2015 International Journal of Aeronautical and Space Sc Vol.16 No.3

Experimental and numerical investigations were conducted to identify the wake characteristics downstream of two vanetype vortex generators over laminar flat plate boundary layer. Experimental study was carried out by using the stereoscopic particle image velocimetry. To describe the flow field around the vortex generator in detail, numerical study was performed. We considered two different planform shapes of vortex generator: triangular and rectangular shape. The height of the generator was chosen to be about the boundary layer thickness at the position of its installation. Two different lengths of the generator were chosen: two and five times the height. Wake measurements were carried out at three angles of attack for each configuration. Wake characteristics for each case such as overall vortical structure, vorticity distribution, and location of vortex center with downstream distance were obtained from the PIV data. Wake characteristics, as expected, were found to vary strongly with the geometry and angle of attack so that no general tendency could be deduced. Causes of this irregular tendency were explained by using the results of the numerical simulation.

[7]

HoJoon YOOK 한국색채학회 2021 International Invitation Exhibition of Color Works Vol.- No.-