DOI QR코드

DOI QR Code

Analysis of the Unstructured Traffic Report from Traffic Broadcasting Network by Adapting the Text Mining Methodology

텍스트 마이닝을 적용한 한국교통방송제보 비정형데이터의 분석

  • 노유진 (도로교통공단, 부경대학교 공간정보시스템공학과) ;
  • 배상훈 (부경대학교 공간정보시스템공학과)
  • Received : 2018.04.16
  • Accepted : 2018.05.31
  • Published : 2018.06.30

Abstract

The traffic accident reports that are generated by the Traffic Broadcasting Networks(TBN) are unstructured data. It, however, has the value as some sort of real-time traffic information generated by the viewpoint of the drives and/or pedestrians that were on the roads, the time and spots, not the offender or the victim who caused the traffic accidents. However, the traffic accident reports, which are big data, were not applied to traffic accident analysis and traffic related research commonly. This study adopting text-mining technique was able to provide a clue for utilizing it for the impacts of traffic accidents. Seven years of traffic reports were grasped by this analysis. By analyzing the reports, it was possible to identify the road names, accident spot names, time, and to identify factors that have the greatest influence on other drivers due to traffic accidents. Authors plan to combine unstructured accident data with traffic reports for further study.

교통사고 관련 제보는 비정형 데이터로서 교통사고를 유발한 가해자나 피해자의 관점이 아닌, 교통사고 발생 지점과 구간, 시간대에 있었던 타 운전자의 관점에서 생성된 교통정보의 가치를 가지고 있다. 그러나, 비정형 데이터인 교통제보가 빅 데이터로서 교통사고 통계나 교통관련 연구에 활용되지 못하였으나, 텍스트 마이닝 기법을 활용한 본 연구를 통해 비정형의 빅 데이터를 시각화하고 해석하여, 기존의 정형 데이터에서 분석하지 못한 정보를 도출할 수 있었다. 그리고 교통사고 발생으로 인한 도로상 영향을 파악할 수 있었다. 이러한 분석으로 교통제보의 트랜드를 파악하고, 운전자가 제보하는 "도로명", "지점명", "시간대"를 추출하였으며, 교통사고 발생으로 다른 운전자에게 가장 많은 영향을 미치는 지점과 구간의 파악이 가능하였다. 향후 실제 교통사고 데이터와 결합하여 교통제보와의 상관성 분석 등을 통해 비정형 데이터의 활용방안을 모색할 계획이다.

Keywords

References

  1. Ahn S. and Cho S.(2010), "Stock prediction using news text mining and time series analysis," In 2010 Conference Proceedings of Korean Institute of Information Scientists and Engineers, 37, pp.364-369.
  2. Bae S. and Park C.(2003), "A Study on the Application of Text Mining to the Analysis of Technical Information," Korea Technology Innovation Society, pp.79-83.
  3. Chen P., Ponocko J., Milosevic N., Nenadic G. and Milosevic J.(2016), "Towards application of text mining for enhanced power network data analytics-part i; retrieval and ranking of textual data from the internet," Mediterranean Conference on Power Generation, Transmission Distribution and Energy Conversion (medpower 2016), pp.1-8.
  4. Choi J., Han H., Lee M. and Ahn J.(2015), "The prediction of Corporate Bankruptcy Using text-mining Methodology," Productivity Review, vol. 29, no. 1, pp.203-206.
  5. Choi Y. and Park S.(2002), "Interplay of Text Mining and Data Mining for Classifying Web Contents," Korean Journal of cognitive science, vol. 13, no. 3, pp.33-35.
  6. Jung C. W.(2016), "A Study on Traffic Accident Investigation Satisfaction Factors," Journal of Transport Research, vol. 23, no. 4, pp.73-84.
  7. Kim K. and Oh S.(2009), "Methodology for Applying Text Mining Techniques to Analyzing Online Customer Reviews for Market Segmentation," The Journal of the Korea Contents Association, vol. 9, no. 8, pp.272-284. https://doi.org/10.5392/JKCA.2009.9.8.272
  8. Kim Y., Heo J. and Kang K.(2015), "Overview of cargo accident using text mining," 2015 Conference of Korea Transportation Research Society, pp.338-343.
  9. Lee K., Roh Y., Yoon S. and Cho Y.(2014), "Structuring of unstructured big data and visual interpretation," Journal of the Korean & Information Science Society, vol. 25, no. 6, pp.1436-1437.
  10. Lee Y., Lim C., Heo M. and Kim H.(2016), "Text-mining technique for Weather call center data analysis," In 2016 Spring Conference Proceedings of Korean Meteorological Society, pp.153-154.
  11. Sun H., Lim C. and Lee Y.(2017), "Analysis of the Yearbook from the korea Meteorological Administration Using a text-mining agorithm," The Korean Journal of Applied Statistics, vol. 30, no. 4, pp.603-613. https://doi.org/10.5351/KJAS.2017.30.4.603