본문 바로가기
KERD
논문KCI 등재

머신러닝 모델을 활용한 한국과 중국 전통 백자의 생산국가 판별 연구: 백자의 화학적 분석 데이터를 기반으로

A Study on the Classification of the Production Country of Korean and Chinese Traditional White Porcelain Using Machine Learning Models: Based on the Chemical Analysis Data of White Porcelain

안상두|조예은
한국문화재보존과학회 (The Korean Society Of Conservation Science For Cultural Properties)|2024.06|보존과학회지 (Journal of Conservation Science)|제40권 제2호|pp. 102-112|DOI 10.12654/JCS.2024.40.2.01

1

조회

0

다운로드

0

인용

0

참고문헌

키워드

Machine learning머신러닝백자White porcelain미량원소지도학습Trace elementsSupervised learning희토류원소생산지 판별Provenance discriminationRare earth elements
주제분류자연과학일반

초록

본 연구의 목적은 도토기 생산지 판별에 있어서 머신러닝 기법의 적용 가능성을 확인하는 것이다. 한국과 중국의 전통 백자 태토의 화학성분을 분석한 데이터 732개를 수집하였고,다양한 머신러닝 알고리즘을 적용하여 생산국가 판별 모델을 개발하였다. 개발된 모델을 146개테스트 시료에 적용한 결과, 통계적 분석법인 principal component analysis-linear discriminantanalysis 모델에서는 87.7%의 예측 정확도를 얻었지만, 머신러닝 기법인 decision tree, K-nearestneighbor, support vector machine 모델에서는 각각 96.6%, 98.6%, 99.3%의 비교적 높은 예측 정확도를 얻을 수 있었다. 아울러 특성 중요도 분석을 통해, 복잡한 데이터 구조의 분류 성능이우수한 머신러닝 기법에서 공통적으로 rubidium을 생산지 판별에 가장 중요한 변수로 활용하고있음을 확인하였다. 이러한 결과는 머신러닝 기법이 한국과 중국의 백자 생산지 판별에 효과적으로 활용될 수 있음을 보여주었다.

Abstract

The objective of this study is to investigate the applicability of machine learningtechniques in discriminating the production origins of pottery. 732 data sets analyzing the chemicalcomposition of traditional white porcelain from Korea and China were collected, and models fordetermining the production country were developed by applying various machine learning algorithms. Upon applying these models to 146 test samples, the statistical analysis, principal componentanalysis-line discriminant analysis yielded a prediction accuracy of 87.7%, while machine learningtechniques such as decision tree, K-nearest neighbor, and support vector machine models demonstratedrelatively high prediction accuracies of 96.6%, 98.6%, and 99.3%, respectively. Additionally, featureimportance analysis confirmed that rubidium is consistently the most critical variable for determiningthe origin in machine learning techniques that exhibit superior performance in classifying complexdata structures. These findings underscore the potential of machine learning techniques in effectivelydiscerning the production origins of white porcelain from Korea and China.

참고문헌 (53)

  1. Abu Alfeilat, H. A., "Effects of Distance Measure Choice on K-Nearest Neighbor Classifier Performance – A Review", Big Data, 7(4), pp.221~248 (2019)
  2. Ahn, S. Y., "Study of Material Characteristics by a Componential Analysis on the Whiteware from the Kiln of Chiljeon-ri, Bangsan-myeon, Yanggu-gun", Journal of Conservation Science, 29(3), pp.261~277 (2013)
  3. Anglisano, A., "Application of Supervised Machine – Learning Methods for Attesting Provenance in Catalan Traditional Pottery Industry", Minerals, 10(1), pp.8~ (2020)
  4. Bartle, E. K., "Provenance determination of oriental porcelain using laser ablationinductively coupled plasma-mass spectrometry(LA-ICPMS)", Journal of Forensic Sciences, 52(2), pp.341~348 (2007)
  5. Breiman, L., "Classification and Regression Trees", pp.368~ (1984), Routledge
  6. Bzdok, D., "Statistics versus Machine Learning", Nature Methods, 15, pp.233~234 (2018)
  7. Charbuty, B., "Classification based on decision tree algorithm for machine learning", Journal of Applied Science and Technology Trends, 2(1), pp.20~28 (2021)
  8. Chen, T., "Provenance studies of the earliest Chinese protoporcelain using instrumental neutron activation analysis", Journal of Archaeological Science, 26(8), pp.1003~1015 (1999)
  9. Kiln site of Bangsan-dong : report on archeological excavation of Celadon & Koryo white porcelain kiln site at Bangsan-dong, Siheung-si, Kyonggi-do, "Choi, K.", Haegang Ceramics Museum, 343~381 (2001)
  10. Cortes, C., "Support-vector networks", Machine Learning, 20, pp.273~297 (1995)
  11. Dias, M. I., "Tracing the origin of blue and white Chinese Porcelain ordered for the Portuguese market during the Ming dynasty using INAA", Journal of Archaeological Science, 40(7), pp.3046~3057 (2013)
  12. Fischer, C., "Export Chinese blue-and-white porcelain : compositional analysis and sourcing using non-invasive portable XRF and reflectance spectroscopy", Journal of Archaeological Science, 80, pp.14~26 (2017)
  13. Gang, G. I., "Whiteware from the Official Kiln of Choseon Dynasty Period in Beoncheon-ri, Gwangju", Journal of Conservation Science, 12(1), pp.1~14 (2003)
  14. White Porcelain Kiln Site In Hapan-ri, Gapyeong, "Gyeonggi Ceramic Museum", Gyeonggi Ceramic Museum, 260~286 (2007)
  15. White Porcelain Kiln Site No.18 in Sindae-ri, Gwangju, "Gyeonggi Ceramic Museum", Gyeonggi Ceramic Museum, 148~155 (2008)
  16. Ceramic Culture, "Gyeonggi Ceramic Museum", Gyeonggi Ceramic Museum, 61~88 (2009)
  17. White porcelain kiln sites no. 8 in Buncheon-ri and no. 2 in Sundong-ri, Gwangju-si, "Gyeonggi Ceramic Museum", Gyeonggi Ceramic Museum, 310~346 (2013)
  18. Analytical Report of the Royal Kiln Complex at Gwangju in Gyeonggi Province, "Gyeonggi Provincial Museum", Gyeonggi Ceramic Museum, 16~277 (2008)
  19. Han, M. S., "Scientific provenance study for Goryeo celadon excavated from seabed", Ph.D., pp.2006
  20. Kang, H. T., "A Study on the Classification of Korean Procelainshards by Neutron Activation Analysis", Bulletin of Environmental Sciences, 10, pp.29~35 (1989)
  21. Katsuki, H., "Comparative study on old white porcelains produced in Gwangju(Korea)and Arita(Japan)in the 15th to 19th centuries", Journal of Archaeological Science : Reports, 27, pp.101927~ (2019)
  22. Laaksonen, J., "Classification with learning k-nearest neighbors", Proceedings of International Conference on Neural Networks(ICNN’96), pp.1480~1483 (1996)
  23. Lee, B. H., "A Study on Chemical Composition and Firing Temperature of White Wares from the Guyre-2ri Kiln Sites, Wonju", Journal of Conservation Science, 30(3), pp.249~261 (2014)
  24. Lee, B. H., "Material Characteristics of White Wares from Yeongdong Province, Gangwon-do – Gangneung and Donghae Kiln Sites", Journal of Conservation Science, 31(3), pp.181~192 (2015)
  25. Lee, B. H., "A Study on the Relationships between White Wares and Raw Materials Excavated at Goseong-ri Kiln Site in Chuncheon, Yeongseo Province, Gangwon-do", Journal of Conservation Science, 36(3), pp.152~161 (2020)
  26. Lerman, P. M., "Fitting Segmented Regression Models by Grid Search", Applied Statistics, 29(1), pp.77~ (1980)
  27. Li, B. P., "ICP-MS trace element analysis of Song dynasty porcelains from Ding, Jiexiu and Guantai kilns, north China", Journal of Archaeological Science, 32(2), pp.251~259 (2005)
  28. Li, B. P., "Application of ICP-MS trace element analysis in study of ancient Chinese ceramics", Chinese Science Bulletin, 48, pp.1219~1224 (2003)
  29. Li, Z., "Rediscovering the largest kiln site in the middle Yangtze River Valley – Insights into Qingbai and grey-greenish ware production at Husi kiln site based on bulk chemical analysis", Archaeological and Anthropological Sciences, 13, pp.218~ (2021)
  30. Lundberg, S. M., "A unified approach to interpreting model predictions", Advances in Neural Information Processing Systems, 30, pp.4765~4774 (2017)
  31. Ma, H., "Provenance of Zhangzhou export blue-and-white and its clay source", Journal of Archaeological Science, 39(5), pp.1218~1226 (2012)
  32. Maione, C., "Predicting the botanical and geographical origin of honey with multivariate data analysis and machine learning techniques – A review", Computers and Electronics in Agriculture, 157, pp.436~446 (2019)
  33. Nembrini, S., "The revival of the Gini importance?", Bioinformatics, 34(21), pp.3711~3718 (2018)
  34. Niziolek, L. C., "A compositional study of a selection of Song Dynasty Chinese ceramics from the Java Sea Shipwreck – Results from LA-ICP-MS analysis", Journal of Indo-Pacific Archaeology, 35, pp.48~66 (2015)
  35. Pal, M., "An assessment of the effectiveness of decision tree methods for land cover classification", Remote Sensing of Environment, 86(4), pp.554~565 (2003)
  36. Report on the Excavation of Gatjom-gol Jeonggak-ri site, Puyo, "Park, H. S.", Chungcheong Institute of Cultural Heritage, 159~178 (2002)
  37. Patle, A., "SVM kernel functions for classification", 2013 International Conference on Advances in Technology and Engineering (ICATE), pp.1~9 (2013)
  38. Pedregosa, F., "Scikit-learn – Machine Learning in Python", Journal of Machine Learning Research, 12, pp.2825~2830 (2011)
  39. Raju, V. G., "Study the influence of normalization", transformation process on the accuracy of supervised classification, 2020 (2020), 729~735
  40. Ruschioni, G., "Supervised learning algorithms as a tool for archaeology – Classification of ceramic samples described by chemical element concentrations", Journal of Archaeological Science : Reports, 49, pp.103995~ (2023)
  41. Ryu, J. M., "A Study on Guidance of Understanding about Chosun White Porcelain", Journal of Art Education, 15, pp.33~57 (2004)
  42. Seoul Museum of History, "The Kiln Sites and Ceramic of Seoul", 250~277 (2006)
  43. Simsek, G., "Toward a fast non-destructive identification of pottery – The sourcing of 14th–16th century Vietnamese and Chinese ceramic shards", Journal of Cultural Heritage, 16(2), pp.159~172 (2015)
  44. Sun, H., "A new classification method of ancient Chinese ceramics based on machine learning and component analysis", Ceramics International, 46(6), pp.8104~8110 (2020)
  45. Tharwat, A., "Principal component analysis-a tutorial", International Journal of Applied Pattern Recognition, 3(3), pp.197~240 (2016)
  46. Trindade, M. J., "Bromine volatilization during firing of calcareous and non-calcareous clays – Archaeometric implications", Applied Clay Science, 53, pp.489~499 (2011)
  47. Tripati, S., "Chemical composition and provenance of Chinese porcelain shards recovered from Old Goa, west coast of India", Journal of Archaeological Science : Reports, 14, pp.467~478 (2017)
  48. Wang, M., "Provenance of Qingbai wares from Shabian kiln archaeological site in the Northern Song dynasty(960–1127 ce)using XRF and LA-ICP-MS", Archaeometry, 62(3), pp.538~549 (2020)
  49. Wilke, D., "Is Non-destructive Provenancing of Pottery Possible With Just a Few Discriminative Trace Elements?", STAR : Science & Technology of Archaeological Research, 2(2), pp.141~158 (2016)
  50. Xu, W., "Compositional analysis below the production region level – A case study of porcelain production at Dehua, Fujian, China", Journal of Archaeological Science, 135, pp.105481~ (2021)
  51. Yap, C. T., "Non-destructive Spectrometric Determination of Trace Element Concentration of Rubidium, Strontium, Yttrium Zirconium and Niobium in Ceramics", Zeitschrift für Naturforschung A, 42(11), pp.1253~1256 (1987)
  52. Zhao, H., "Multi-element composition of wheat grain and provenance soil and their potentialities as fingerprints of geographical origin", Journal of Cereal Science, 57(3), pp.391~397 (2013)
  53. Zhu, T., "Research on the Characteristics of Chemical Elements in Chinese Ancient Bluish-white Porcelains from Different Kilns", Rock and Mineral Analysis, 25(2), pp.114~118 (2006)

관련 논문AI추천

중성자방사화분석을 활용한 경상남도 백자의 산지 및 특성 분류김규호, 김나영한국문화재보존과학회 · 보존과학회지2007 · 제21권 강원 영서지역 춘천 고성리 가마터 출토 백자 및 태토 원료의 상관관계 연구이병훈한국문화재보존과학회 · 보존과학회지2020 · 제36권 제3호A Study of the Chemical Composition of Korean Traditional Ceramics (I): Celadon and Koryŏ Whiteware고경신, 김규호, 안상두, 이연숙, 이영은, 주웅길한국문화재보존과학회 · 보존과학회지2010 · 제26권 제3호A Study of the Chemical Composition of Korean Traditional Ceramics (III): Comparison of Punch’ŏng with Koryŏ Ware and Chosŏn Whiteware고경신, 김규호, 안상두, 이연숙, 이영은, 주웅길한국문화재보존과학회 · 보존과학회지2011 · 제27권 제1호Changes of Physical and Chemical Properties for Making Raw Materials and Reproductions According to Manufacturing Stages in Traditional Korean White Porcelain김두현, 오은정, 정지윤, 한민수한국문화재보존과학회 · 보존과학회지2022 · 제38권 제4호김제 청도리 동곡마을 도요지 출토 도자기의 고고화학적 특성 분석김규호, 김나영, 박영아, 전유리한국문화재보존과학회 · 보존과학회지2017 · 제33권 제2호

이 간행물의 다른 논문 (7건)

중성자 분석을 이용한 철제유물의 부식시스템 연구김이연, 김종열, 조남철, 김태주pp. 113-122남원 선원사 목조지장보살삼존상 및 소조시왕상의 채색안료 분석과 색상 다양성 해석정동희, 한두루, 이찬희, 최일규pp. 123-137가죽 문화재의 유연성 복원을 위한 표면 처리제 연구: 표면 처리된 가죽의 물리적특성을 중심으로유수정, 박지선pp. 138-1476⋅25전사자 故김기봉 이등중사 유품의 과학적 보존처리위광철, 이호연pp. 148-156라오스 참파삭 홍낭시다 사원 출토 고대 녹색 유리의 화학조성과 의미최혜지, 양혜리, 이찬희, 전유근pp. 157-167
원문 보기