An Efficient Analysis of Road Accident using Text Mining

An Efficient Analysis of Road Accident using Text Mining

Mrs. C. Nithya

Assistant Professor

Department of Computer Science and Engineering. Muthayammal Engineering College (Autonomous) Rasipuram.

P. Hari Prasath

UG Scholar

Department of Computer Science and Engineering. Muthayammal Engineering College (Autonomous) Rasipuram.

Abstract:- There is nothing more valuable than a human life in the world. Road accidents has killed more people than anything else in the world. Thereare millions of people being killed by road accidents around the world every year . n nIn this dissertation work, the focus is on the causes of roadaccidents as to what are the reasons that cause these dreadful accidents and how can we lead to a solution which can be limit or minimizethese accidents. In this work, we have taken thousands of road accidents records which will be useful to understanding the major causes ofroad accidents and how those accidents can be resolved. There are various cause because of which the accidents are caused. From Overspeeding to Drink and Drive, we have tried to cover all the causes which are somehow involved in road accidents. To understand the patternsof road accidents we need Data and to gain some useful information from Datasets there is a concept called as Data Mining.

Keywords Facial Expression Recognition (FER), multimodal sensor data, emotional expression recognition, spontaneous expression, real-world conditions

1. INTRODUCTION

   This dissertation work is based on Data Mining domain. In Data mining, the large amount of dataset is taken (more than 5 years, 10 ye are of data) and we try to uncover some useful and hidden information from the data is made free from noise. This process is called as pre- processing of the data. Now the data is free from all the inconsistencies. Next step is to transform data into a common format which makes it an appropriate form for applying data analytics and data mining. After this step, we apply different data mining techniques such as Classification, Clustering and Rule Association Mining to extract hidden and useful information from the huge dataset. Once we have the hidden information, we can visualize the same using Pie- chart, Bar-chart, etc. The whole process is also termed as Knowledge Discovery.

   The main objective of this application is to provide a solution for accident detection and prevention of human life safety. The application has been divided into four modules based on functionalities.

   K. Jagareesh

   UG Scholar

   Department of Computer Science and Engineering. Muthayammal Engineering College (Autonomous) Rasipuram

   The module is designed to build up an integrated system to cover various aspects of android based Automatic vehicle Accident Detection by Using Android application. The application is designed using Road Accident Using Text Mining.

2. LITERATURE REVIEW

   The authors have performed an analysis on road accidents in India. Different classification algorithms have been applied on Open Government dataset to classify accident causes into different classes. The limitation in this work is that no Training has been provided to the system where it can make predictions with any new dataset.

   The authors have performed research work on causes of road accident.

   No training or testing has been done on real time data. Causes Support Model have been applied to understand causes of road accidents. There are different classification algorithms such as K-NN, Naive Bayes, which are not been applied to understand the real causes of accidents. No performance parameters have been applied to check how well and correct the system is performing.

   In this work, the authors have done a research work on road accidents using different data mining techniques . Only survey has been done and no implementation has been performed using any data mining tool. K- means Clustering has been applied to cluster different causes of accidents. No real time data has been used. Also, the accuracy achieved is low.

   In this paper, the authors have applied correlation analysis and exploratory visualization techniques to understand the causes of road accidents in Rajasthan State of India . In this paper, only clustering has been applied and no real causes has been uncovered for road accidents. The authors have focused on the type of road in State highways and districts which have been concluded to be in bad condition and

   being the causes for road accidents in state. There have been no performance measures taken in this research work.

3. ANALYSIS AND DESIGN OF THE APPLICATION

   A.E XISTING WORK

   Using appropriate system methods to improve safety measures is an remarkable challenge, particularly when the sum of existing records often increases. From the present circumstance, you can choose some existing studies that mainly address local traffic accident data. Iraq is one of the nations with high traffic accident deaths and casualties. In the past 3 years, traffic accidents in Iraq resulted in an average of 24,000 deaths per day (three people per hour), and about 240,000 people were injured every year. This fact prompted a group of writers to classify the most noteworthy affecting the severity of driver injury in these road traffic accidents. Not able to detect the accident, so we cant able to save the people quicklyDeath rate might increase because of no technique to detect the accidentVillage area and remote area people are heavily affected.

   1. Proposed Work:

      The proposed model for accident zone detection system can prove to be an important aid in constructing smart transport systems in near future if implemented properly Also the system can be used by the owners of the transport companies etc to monitor the accident location using the web app. These features can also help in case of vehicle theft etc.

      It can also overcome the issue of lack of automated system for the detection of the site of accident. It is very useful to avoid the accident easily

   2. Advantages:

   • Easily find the accident location with the help of Data Mining

   • Treatment also get early, so we can save the lives as possible as soon.

4. SYSTEM IMPLEMENTATION

   1. User: User can view the accident zone by giving the travel location. If the user give the travel route, I.e. source place and destination place, the system shows the accident zone between source and destination places.

   2. Admin: Admin can handle the over all process of this application, through this module, the admin can add the locations. The admin can input the data sets, that means, admin can add the accidents details by giving the route details and accident happened place.

   3. Generate Accident Report: In this module the admin can generate the accident report. Those reports are categorize by accident zone and dangerous zone. Accident zone is a place of accidents happens few time and few deaths and injure. Dangerous zone is a place of accident happens many times and many deaths and injures.

   4. Characteristics of Accident Report: The Characteristics of accident reports are categorize by accident zone and dangerous zone. Accident zone is a place of accidents happens few time and few deaths and injure. Dangerous zone is a place of accident happens many times and many deaths and injures.

   5. Stored in databases: All data are stored in the data base previously.

5. ARCHITECTURE DESIGN

6. RESULT & DISCUSSION

   It can be seen tht the largest number of accidents occurred in uncontrolled areas, causing 128,263 accidents in traffic control police control areas, and the number of reported accidents was 1,66,158.

   The below figure shows the pie chart of the accidents in the urban as well as rural area where the road is controlled by traffic or police.

   Accidents in controlled area Accidents based on the Age of Vehicles

   The below table and the corresponding figure shows the data regarding the age of vehicles involved in the accidents. During the year, fewer than five-year-old vehicles had the highest number of accidents in the country (3,94,198), of which 56,329 were dead and 2,030,042 were injured.

   Graphical representation of age of vehicles

   Persons killed in Road Accidents in terms of Road User Categories:

   The road is used by two categories namely vulnerable road users who are largely unprotected like pedestrian bicycle riders and two wheeler riders. The second category are the drivers.

   The above table and the figure shows that the major causality is affected to the vulnerable road users since they are the one who are more exposed. They have the highest tendency towards accidents.

   Accident on based on Type of vehicle

7. CONCLUSION

   From the statistical results, it can be seen that the rural mortality rate is higher, while the city is lower. Statistical analysis also includes other limiting factors such as the age of the vehicle, the type of vehicle, the age group of the person, and the category of road users. The predicted data results are displayed in a graphical

   representation. Graphical representations help the public understand accident metrics that help reduce mortality.

8. REFERENCES

1. F.M.O.I. Forensic Medicine Organization of Iran; Statistical Data, Accidents, online avail- ableon: http://www.lmo.ir/?siteid=1&pageid=1347

2. L.Y. Chang, H.W. Wang, Analysis of traffic injury severity: An application of non-parametric classification tree techniques,

   Accident Analysis and Prevention, 38(5), pp. 1019-1027, 2006

3. R. Nayak et al., Road Crash Proneness Prediction using Data Mining. Ailamaki, Anastasia & Amer-Yahia , Sihem (Eds.) Proceedings of the 14th International Conference on Extending Database Technology, Association for Computing Machinery (ACM), Upp-sala, Sweden, pp. 521-526, 2011.

4. S. Vigneswaran et al., Efficient Analysis of Traffic Accident Using Mining Techniques, International Journal of Software and Hardware Research in Engineering, Vol. 2, No. 3, 2014, pp. 110- 118, 2014.

5. L. Martin et al. Using data mining techniques to road safety improvement in Spanish roads, XI Congreso de Ingeniería del Transporte (CIT 2014), Procedia – Social and Behavioral Sciences 160 (2014), pp. 607614, 2014

6. Gurubhagavatula, I., Nkwuo, J. E., Maislin, G., and Pack, A. I. (2008). Estimated cost of crashes screening and treatment of obstructive sleep apnea. International Journal of Accident Analysis & Prevention, 40 (1), 104-115.

7. Chatterjee, S. (1998). A connectionist approach for classifying accident narratives. Purdue University.

8. Li-Yen, C., & Wen-Chieh, C. (2005) Data mining of tree-based models to analyze freeway accident frequency.