道路交通事故的發生和道路交通量的增加密切相關。而交通事故一直是一個嚴重的社會問題，政府部門無不大力提倡交通安全議題。然而，還是有許多改進的空間。如今的數據科學，運用許多分析技能在許多領域來探勘事件造成因素、模式或事實。這種科學方法當然也可以協助我們預防和減少交通事故，但在此之前，數據本身的質量才是數據分析的基礎。本研究的目的是幫助花蓮縣交通警察隊在建立交通事故數據分析時，能有一套詳細且標準的處理指導方針。此指南建立於本研究的數據檢索系統和數據管理系統，為的是使數據預處理階段能夠自動化進行，而不會在減少操作時間的同時降低數據的準確性。在高質量數據的前提下，數據分析的結果將具有更真實的價值和參考意義。

The occurrence of road traffic accidents and the increase of road traffic are closely related. Traffic accidents have always been a serious social problem. All government departments vigorously promote traffic safety issues, however, there is still a lot of room for improvement. Today's data science uses many analytical skills
to explore events, incident patterns or facts in many areas. This scientific method can also help us to prevent and reduce traffic accidents, but it requires a good quality of the data because itself is the basis for data analysis. The purpose of this study is to help Hualien County Traffic Police Team to maintain a set of detailed and
standard guidelines for handling traffic accident data analysis.This guide is based on the data retrieval system and data management system of this study. In order to automate the data pre-processing phase without reducing the accuracy of the data while reducing the operating time. With the premise of high-quality data, the results
of data analysis will have more real value and reference value.

Chapter 1 Introduction 1
1.1 Background and Motivation 1
1.2 Research Purpose 4
1.3 Subjects 5
1.4 Research Method and Design 6
1.4.1 Research Method 6
1.4.2 Research Design 7
Chapter 2 Literature Review 11
2.1 Definition and classification of traffic accidents 11
2.1.1 Definition of road traffic accident 11
2.1.2 Classification of road traffic accidents 12
2.2 Road Traffic Accident Investigation Literatures 13
2.3 Road Traffic Accident Electronic Management 13
2.4 Data interpretation and Data administration 14
2.5 Road Traffic Accident data statistic and analysis 17
Chapter 3 Data Archiving and Management Process 21
3.1 Traffic accident data sources and filing methods 21
3.1.1 Types of traffic accidents related forms 21
3.1.2 Description of Road Traffic Accident Questionnaire 23
3.1.3 The current road traffic accident data e-file mode 31
3.1.4 The feedback of operation problems of current e-system data archiving user 32
3.2 Traffic accident archiving data management system 33
3.2.1 Classification of traffic accident data content 33
3.2.1 Traffic accident data problems and recommendations 35
3.2.2 Data administrative structure 42
3.2.3 Application 49
Chapter 4 Traffic Accident Data Analysis Results 51
4.1 Traffic Accident Statistical Indicators in Hualien County 51
4.1.1 Numbers of traffic accidents and casualties in category A1, A2 and A3 from 2014 to 2016 51
4.1.2 Category A1A2 Traffic Accident Statistics in Hualien County from 2014 to 2016 53
4.1.3 Category A1A2 incident number and casualty monthly statistics in Hualian County 53
4.1.4 Monthly overview of each area 59
4.1.5 Comparison of the probability of accident in Hualien County Administrative Area 62
4.1.6 Comparative analysis of category A1 and A2 incident number in Hualien County 63
4.1.7 Comparative analysis of the number of perpetrators in category A1A2 in Hualien County Administrative Region 64
4.1.8 Comparative Analysis of Traffic Accident Mortality in Hualien County Administrative Region 66
4.2 Define Key Groups 68
4.2.1 Group by Age 68
4.2.2 Group by Incident parties distinguish 70
4.2.3 Group by Driving Qualifications 75
4.2.4 Group by Drinking Situation 76
4.3 Key Groups Case Study 81
4.3.1 Ordinary heavy-duty locomotives 82
4.3.2 Own-use passenger vehicle 90
Chapter 5 Conclusion and Suggestion 99
5.1 Conclusion 99
5.2 Suggestion for the Further Research 99
References 101

1. 廖銘亮(民93)，「建立先進交通事故資料管理系統雛型之研究─以台中市交通事故資料為例」，逢甲大學交通工程與管理學系碩士班碩士論文。
2. 內政部警政署、中央警察大學合編（民96），「道路交通事故現場圖與調查報告表填表須知」，未出版。
3. 歐俐伶、楊東謀(民105)，「台灣政府開放資料之詮釋資料建置探討」，教育資料與圖書館學　53 : 1 (Winter 2016)。
4. 周文生、陳雯龍(民91)，「交通事故調查資料作業流程之探討」，道路交通安全與執法研討會。
5. 王蓓瑜(民98)，「美術館數位典藏詮釋資料欄位專業使用者需求之研究」，國立台灣師大學圖書資訊學研究所碩士學位論文。
6. 鐘佳霖(民105)，「建構資料欄位關係之知識本體-以空氣品質、國土利用、人口統計資料為例」，國立成功大學都市計畫研究所碩士論文。
7. 王秀雯(民93)，「應用資料挖掘技術於交通事故傷亡嚴重程度之研究」，國立嘉義大學運輸與物流工程研究所碩士論文。
8. 林郁志(民86)，「都市地區肇事嚴重程度之分析研究 -以臺南市為例」，國立成功大學交通管理學系碩士論文。
9. 謝鍚釗(民89)，「道路交通安全改善計畫績效評估程序之研究」，中央警察大學交通管理研究所碩士論文。
10. 陳鴻斌(民91)，「道路交通事故調查表報之研究」，中央警察大學交通管理研究所碩士論文。
11. 葉名山、廖偉呈與胡振華(民96)，「警員交通事故管理系統之建置」，九十六年道路交通安全與執法研討會。
12. 郭國津(民93)，「道路交通事故資料庫管理系統之建立」，中央警察大學交通管理研究所碩士論文。
13. 王仲麒(民102)，「Excel 2013商業智慧分析:資料處理x樞紐分析x Big data分析工具PowerPivot及PowerView」，峰出版社。
14. (意)法拉利(民104)，「微軟Excel 2013:用PowerPivot建立數據模型」，清華大學出版社。
15. 謝邦昌(民104)，「資料採礦完全手冊」，中華資料採礦協會。
16. 謝邦昌、研究團隊、蔣以仁、李毅、謝邦彥(民104)，「Text Mining文本探勘」，中華資料採礦協會。
17. 謝邦昌、鄭宇庭、宋龍華、陳妙華(民105)，「大數據分析Excel Power BI全方位應用」，碁峰資訊股份有限公司。
18. Abdelwahab, H. T.and Abdel-Aty, M. A., (2001). “Development of Artificial Neural Network Models to Predict Driver Injury Severity in Traffic Accidents at Signalizes Intersection”. Transportation Research Record. 1746, Paper No. 01-2234, pp. 6-13.
19. Bedard, M., Guyatt, G., Stones, M. and Hirdes, J. (2002). “The Independent Contribution of Driver, Crash and Vehicle Characteristics to Driver Fatalities”. Accident Analysis and Prevention, vol.34, No. 6, pp. 717-727.
20. Kockelman, K. M. and M. and Kweon, Y.J. (2002). “Driver Injury Severity: An Application of Ordered Probit Models”. Accident Analysis and Prevention, vol.34, No. 3, pp. 313-321.
21. Mussone, L., Ferrari, A and Oneta, M. (1999). “An Analysis of Urban Collisions Using an Artificial Intelligence Model”. Accident Analysis and Prevention, vol.31, pp. 705-718.
22. Dursun D., Ramesh S. and Max B., 2006. “Identifying Significant Predictors of Injury Severity in Traffic Accidents Using a Series of Artificial Neural Networks”. Accident Analysis and Prevention, vol. 38, No. 3, 2006, pp. 434-444.
23. Farmer, C. M., Braver, E. R. and Mitter, E. L., 1997. “Two-vehicle Side Impact Crashes: The Relationship of Vehicle and Crash Characteristics to Injury Severity”. Accident Analysis and Prevention, vol. 29, No. 3, pp. 399-406.