Türk Karasularında Meydana Gelen Gemi Kazalarının Analizi: Bir Veri Madenciliği Uygulaması

Dünya ticaret hacminin büyük bir bölümünün taşınmasına aracılık eden denizyolu taşımacılığı, içinde bulunduğu koşulların değişkenliğinden ötürü her an tehlike ile karşılaşılma olasılığı yüksek bir taşımacılık türüdür. Yaşanabilecek en ufak bir olumsuzluğun dahi çok tehlikeli sonuçlar doğurduğu geçmiş yıllarda görülmüştür. Bu sebeple kaza nedenlerini tespit ederek farkındalığı artırmak, önleyici tedbirler geliştirmek için politika oluşturmak adına deniz kaza analizlerinin doğru bir biçimde yapılması ve değerlendirilmesi büyük önem arz etmektedir. Büyük veri yığınları içinden anlamlı bilgilere ulaşıp bilgisayar programlarıyla tahmin edici ve tanımlayıcı yorumlar yapmamıza olanak sağlayan veri madenciliği yöntemiyle deniz kazalarının analizinin yapılması çalışmanın temel konusunu oluşturmaktadır. Bu çalışmada Türk karasularında gerçekleşen deniz kazaları incelenmiştir. Bu bağlamda çalışmanın amacını deniz kazalarında hangi değişkenlerin birlikte hareket ettiğini, veri madenciliğinin önemli analiz yöntemlerinden biri olan birliktelik kuralıyla tespit etmek oluşturmaktadır. Yapılan analizler neticesinde kazalarda; gemi yük durumu, gemide kılavuz kaptanın varlığı, baş iter ve kıç iter gibi donanımların durumu, gemi bayrağı, gemi tipi ve meteorolojik etkenlerin etkili birer değişken olduğu tespit edilmiştir.

Analysis of Ship Accidents in Turkish Territories: A Data Mining Application

Maritime transportation which mediates the transmission of major part of the world’s trading volume is a type of transportation with high probability of encountering dangerous situations due to the instability of its conditions. In the past years, it became clear that the even smallest negativity caused perilous results. Thus, accurate implementation and evaluation of sea accident analysis is important to establish a policy for developing preventive measures and increasing awareness by determining the reason of accident. Analysis of sea accidents forms the fundamental subject of the study with data mining method which allow us to make estimated aand definitive interpretations with computer programs by accessing significant information within large data stacks. In this study sea accidents occurring in Turkish territorial waters have been reviewed. In this context, the purpose of the study consists of determining which factors conduct together in sea accidents by association rule which is one of analysis methods of datamining. As a result of analysis, it has been determined that ship’s loading condition, existence of maritime pilot in the ship, conditions of equipments such as bow thruster and quarterdeck thruster, ship flag and type and meteorological elements have been effective factors on the subject.

___

  • Acharya, T. D., Yoo, K. W. ve Lee, D. H. (2017). GIS-based Spatio-temporal Analysis of Marine Accidents Database in the Coastal Zone of Korea. Journal of Coastal Research. The 2nd International Water Safety Symposium. Special Issue No. 79, pp. 114-118. Coconut Creek (Florida) Akten, N. (2006). Gemi Kazaları: Çevre İçin Ciddi Bir Tehdit. Journal of Black Sea/Mediterranean Environment. 12(1): 269-304.
  • Akyüz, E. (2016). A Marine Accident Analysing Model to Evaluate Potential Operational Causes in Cargo Ships. Safety Science 92: 17–25.
  • Asyalı, E. 2014 sunum Gemi Kazaları Nedenleri ve İnsan Faktörü Ulaştırma, Denizcilik Ve Haberleşme Bakanlığı Deniz Kazalarını Araştırma Ve İnceleme Çalıştayı (18–19 OCAK 2014/ANTALYA) http://www.kugm.gov.tr/ BLSM_WIYS/ KAIK/tr/Doc/ 20140228_095711_ 76347_1_64.pdf. Aydogdu, Y. D. (2013). A Comparison of Maritime Risk Perception and Accident Statistics in the Istanbul Straight. The Journal Of Navıgatıon 67: 129–144.
  • Baker, C.C., Seah, A.K. (2004) Maritime Accidents and Human Performance: the Statistical Trail. Martech 2004:225-239
  • Changhai, H. ve Shenping, H. (2018). Factors Correlation Mining On Maritime Accidents Database Using Association Rule Learning Algorithm. Cluster Computing 586- 595.
  • Chauvin, C., Lardjane, S., Morel, G., Clostermann, J. ve Langard, B. (2013). Human and organisational factors in maritime accidents: Analysis of collisions at sea using the HFACS. Accident Analysis and Prevention 59: 26-37.
  • Chen, C., Khoo L. P., Chong, Y. T., Yin, X. F. (2013). Knowledge Discovery Using Genetic Algorithm For Maritime Situational Awareness. Expert Systems With Applications 41: 2742–2753.
  • Dalton, T. ve Jin, D. (2010). Extent and frequency of vessel oil spills in US marine protected areas. Marine Pollution Bulletin 60(2010): 1939–1945.
  • Davies, A. J. ve Hope, M. J. (2015). Bayesian İnference-Based Environmental Decision Support Systems For Oil Spill Response Strategy Selection. Marine Pollution Bulletin 96: 87–102. Dikis, K. ve Lazakis, I. (2019). Dynamic Predictive Reliability Assessment Of Ship Systems. International Journal of Naval Architecture and Ocean Engineering.
  • Dodunekov. S, Minchev, Z., Mitov, I., Ivanova, K.,Dobrinkova, N., Boyvalenkov, P., Pavlov, R. and Kelevedzhiev, E. (2012). Knowledge Dıscovery Methods and Tools and Contınuous Sıtuatıon Awareness Systems (the Bulgarian Academic Approach). Mathematic and Informatics Bulgarian Academy of Science.
  • DTGM (İstatistik Bilgi Sistemi). Deniz Kazaları İstatistikleri 2013-2018. https://atlantis.udhb.gov.tr/istatistik/diger_deniz_kazalari.aspx
  • Ece, N. J. ve Özdemir, Ü. (2017). Kılavuzluk / römorkör servisleri ve teknolojileri 17. Kongresi Kitabı (s.75). İstanbul ve Çanakkale Boğazları’nda Meydana Gelen Deniz Kazalarının Türleri ile Kılavuz Kaptan Alınması Arasındaki İlişkinin Analizi. Düzenleyen Gemi Mühendisleri Odası İzmir. 27-28 Ekim 2017.
  • EMSA (European Maritime Safety Agency). Annual Overview of Marine Casualties and Incidents. 2018.
  • Fayyad, U. (2001). The Digital Phisics of Data Mining. Communications of the ACM 44(3): 62-65.
  • Fricke, W., Cui, W., Kierkegaard, H., Kihl, D., Koval, M., Mikkola, T., Parmentier, G., Toyosada, M., ve Yoon, J. H. (2001). Comparative fatigue strength assessment of a structural detail in a containership using various approaches of classification societies. Marine Structures 15: 1–13.
  • Han, J. ve Kamber, M. (2001). Data Mining: Concepts and Techniques. Burnaby: Morgan Kaufmann Publishers.
  • Hassel, M., Asbjørnslett, B. E. ve Hole, L. P. (2011). Underreporting of maritime accidents to vessel accident databases. Accident Analysis and Prevention 43(2011): 2053– 2063.
  • IMO (International Maritime Organization) (2018). GISIS (Global Integrated Shipping Information System). Marine Casualties and Incidents. http://gisis.imo.org/ Public/MCI/Search.aspx?Mode=Advanced, (08.02.2018).
  • John, A. ve Osue, U. J. (2017). Collision Risk Modelling of Supply Vessels and Offshore Platforms Under Uncertainty. The Journal Of Navıgatıon 70: 870–886.
  • Ketkar, K. W. ve Babu, A. J. G. (1997). An Analysıs of Oil Spills From Vessel Traffic Accidents. Transpn Res.-D, 2(1): 35-41.
  • Kondratenko, Y. P., Kondratenko, G. V., Pidoprigora, D. M., Sidorenko, S. A., ve Timchenko, V. L. (2000). Fuzzy Approach For Desıgn Of Shıp's Decısıon-Makıng Systems. IFAC Management and Control of Production and Logistics 1191-1196.
  • Kouamou, G. (2011). A Software Architecture for Data Mining Environment. Ch.13 in New Fundamental Technologies in Data Mining. InTech Publ., pp.241-258. https://www.intechopen.com/books/new-fundamental-technologies-in-data-mining/a-software-architecture-for-data-mining-environment.
  • Le Blanck, L. A. ve Wyckoff, P. G. (1988). A Strategic Success Factor Analysis of the New Orleans Vessel Traffic Service. Transportation Journal pp.44-50.
  • Le Blanc, L. A. ve Rucks, C.(1996). A Multıple Dıscrımınant Analysıs Of Vessel Accıdents. Accid. Anal. And Prev. 28(4): 501-510.
  • Macrae, C. (2009). Human factors at sea: common patterns of error in groundings and collisions. Maritime Policy & Management 36(1): 21-38.
  • Montewka, J.,Hinz, T., Kujala, P. ve Matusiak, J. (2010). Probability Modelling Of Vessel Collisions. Reliability Engineering and System Safety 95: 573–589.
  • Park, Y. A., Yip, T. L. ve Park, H. G. (2019). An Analysis of Pilotage Marine Accidents in Korea. The Asian Journal of Shipping and Logistics 35(1): 049-054.
  • Raiyan, A., Das, S. ve Islam, M.R. (2017). Event Tree Analysis of Marine Accidents in Bangladesh. Procedia Engineering 194: 276 – 283.
  • Talley, W. K., Jin, D. ve Powell, H. K. (2005). Determinants of crew injuries in vessel accidents. Maritime Policy & Management 32(3): 263-278.
  • Talley, W. K., Jin, D. ve Powell, H. K. (2006). Determinants of severity of passenger vessel accidents. Maritime Policy & Management 33(2): 173-186.
  • Tzannatos, E. ve Kokotos, D. (2009). Analysis of accidents in Greek shipping during the pre- and post-ISM period. Marine Policy 33 (2009): 679–684
  • UAB (Ana Arama Kurtarma Koordinasyon Merkezi). (2017). http://aakkm.udhb.gov .tr/ (23.03.2017)
  • UAB (T.C. Ulaştırma ve Altyapı Bakanlığı). (2018). Deniz Ticareti 2017 İstatistikleri. Ankara: Denizcilik Genel Müdürlüğü.
  • UNCTAD (United Nations Conference on Trade and Development). (2000). Rewiev of Maritime Transport. http://unctad.org/en/Docs/rmt2000_en.pdf. (14.01.2018).
  • UNCTAD (United Nations Conference on Trade and Development). (2017). Rewiev of Maritime Transport. New York.
  • UNCTAD (United Nations Conference on Trade and Development). (2018). Rewiev of Maritime Transport. New York.
  • WEKA (2018). Machine Learning Software in Java.The University of Waikato. https://www.cs.waikato.ac.nz/ml/weka/index.html.
  • Witt, N. A. J., Sutton, R. ve Miller, K. M. (1995). A Track Keeping Neural Network Controller for Ship Guidance. Marine Dynamics Research Group 385-392.
  • Yang, B., Zhao, Z. ve Ma, J. (2018). Marine Accidents Analysis Based on Data Mining Using Kmedoids Clustering and İmproved A Priori Algorithm. IOP Conference Series: Earth and Environmental Science 189: 1-9.