Application of Data Mining Technology to Support Road Traffic Accident Analysis at Addis Ababa Traffic Office

No Thumbnail Available



Journal Title

Journal ISSN

Volume Title


Addis Ababa University


Road traffic accidents are among the top le a din g causes of deaths and injuries of various levels. Ethiopia is one of the countries of the world experiencing highest rate of such accident s resulting in fatalities and various levels of injuries. Addis Ababa, the capital city of Ethiopia, takes the lion 's s hare of the risk having higher number of vehicles and traffic. And the cost of these fatalities and injuries due to such road traffic accidents has a great impact on the socio-economic development of a society. This thesis reports on the finding s of a research that had the objective to build a decision support sys tem to handle road traffic accident analysis, for Addis Ababa C it y Traffic Office. The study focused o n injury severity levels resulting from an accident. In do in g so the aim of this research was to assess the potential applicability of data mining technology specifically decision tree technique to help traffic accident data analysis in decision-making process at the traffic office. In the thesis, the process of building a model through know ledge discovery and data mining techniques on historical accident record data is described. Different tools and techniques are also used for the purpose of data analysis. The methodology adopted had three basic steps name Lydia collection , data preparation, and model building and validation. The required data was selected and extracted from Addis Ababa Traffic Office. Then, data preparation tasks (such as data transformation, deriving of new attributes, and handling o f missing values) were undertaken. The final step was mo del building and validation using the selected tools and techniques. The decision tree Knowledge SEEKER algorithm is used in the stud y. The particular too l used for the mo de l building was the decision tree incorporated in Knowledge STUDIO. After successive experiments, a model that can classify accidents w ell with a better accuracy as fatal , serious, and slight or property-damage was selected and evaluated . Experiment result s reveal that the use o f decision tree is helpful in detecting dangerous accidents through identifying behavioral and roadway accident patterns. The reported findings are promising, making the proposed model a useful" tool in the decision making process. And the whole research process can be a good input for further in-depth research.



Information Science