The Journal of the Convergence on Culture Technology (문화기술의 융합)

The International Promotion Agency of Culture Technology (국제문화기술진흥원)
권호 표지
DOI QR코드

DOI QR Code

Media-based Analysis of Gasoline Inventory with Korean Text Summarization

한국어 문서 요약 기법을 활용한 휘발유 재고량에 대한 미디어 분석

  • 윤성연 (서울여자대학교 데이터사이언스학과) ;
  • 박민서 (서울여자대학교 데이터사이언스학과)
  • Received : 2023.07.11
  • Accepted : 2023.09.05
  • Published : 2023.09.30
Download PDF
⟨ Previous Next ⟩

Abstract

Despite the continued development of alternative energies, fuel consumption is increasing. In particular, the price of gasoline fluctuates greatly according to fluctuations in international oil prices. Gas stations adjust their gasoline inventory to respond to gasoline price fluctuations. In this study, news datasets is used to analyze the gasoline consumption patterns through fluctuations of the gasoline inventory. First, collecting news datasets with web crawling. Second, summarizing news datasets using KoBART, which summarizes the Korean text datasets. Finally, preprocessing and deriving the fluctuations factors through N-Gram Language Model and TF-IDF. Through this study, it is possible to analyze and predict gasoline consumption patterns.

국가 차원의 지속적인 대체 에너지 개발에도 석유 제품의 사용량은 지속적으로 증가하고 있다. 특히, 대표적인 석유 제품인 휘발유는 국제유가의 변동에 그 가격이 크게 변동한다. 주유소에서는 휘발유의 가격 변화에 대응하기 위해 휘발유 재고량을 조절한다. 따라서, 휘발유 재고량의 주요 변화 요인을 분석하여 전반적인 휘발유 소비 행태를 분석할 필요가 있다. 본 연구에서는 주유소의 휘발유 재고량 변화에 영향을 미치는 요인을 파악하기 위해 뉴스 기사를 활용한다. 첫째, 웹 크롤링을 통해 자동으로 휘발유와 관련한 기사를 수집한다. 둘째, 수집한 뉴스 기사를 KoBART(Korean Bidirectional and Auto-Regressive Transformers) 텍스트 요약 모델을 활용하여 요약한다. 셋째, 추출한 요약문을 전처리하고, N-Gram 언어 모델과 TF-IDF(Term Frequency Inverse Document Frequency)를 통해 단어 및 구 단위의 주요 요인을 도출한다. 본 연구를 통해 휘발유 소비 형태의 파악 및 예측이 가능하다.

Keywords

References

  1. Korean Statistical Information Service(KOSIS), Ministry of Trade, Industry and Energy, 2021.
  2. K. S. Cha, "A Study on the Differences of Price Asymmetry and the Price Adjustment Process between Petroleum Products," KUKJE KYUNGJE YONGU, Vol. 26, No. 4, 2020.
  3. D. H. Shin and H. H. Jo, "An Analysis on the Structural Break of Asymmetric Price Effects to Transport Energe Consumption: Evidence from Gasoline and Diesel Consumption in Korea," AKES, Vol. 34, No. 2, 2016.
  4. H. Kim, "An Analysis of the Asymmetry of Domestic Gasoline Price Adjustment to the Crude Oil Price Changes: Using Quantile Autoregressive Distributed Lag Model," Environmental and Resource Economics Review, Vol. 31, No. 4, pp. 755-775, 2022.
  5. K. Cha, "A Study on Price Asymmetry of the Retail Gasoline Market," Journal of Budget and Policy, Vol. 9. No. 4, pp. 31-61, 2020.
  6. J. Bae, S. Kim, M. Kim., S. Oh., and E. Heo, "The Asymmetric Response of Gasoline Prices to International Grude Oil Price Changes Considering Inventories," Envioronmental and Resource Economics Review, Vol. 22, No. 4, pp. 643-670, December 2013. https://doi.org/10.15266/KEREA.2013.22.4.643
  7. H. Jang and B. Choi, "Effects of fuel tax cut on retail prices and its implications," Korean Energy Economic Review, Vol. 22, No. 1, pp. 205-228, March 2023.
  8. B. Seo, "Machine-Learning-Based News Sentiment Index (NSI) of Korea," Bank of Korea WP, Vol. 2022, No. 15, September, 2022.
  9. H. S. Jang, "Effects of temporary tax cut on retail prices: Evidence from Korean gas stations", Korean Energy Economic Review, No. 2, 2021.
  10. A. S. Herbert, "The Choice of a Class Interval." American Statistical Association, Vol. 21, No. 153, pp. 65-66, 1926. https://doi.org/10.1080/01621459.1926.10502161
  11. T. Wang, J. Song, D. Son, M. Kim, D. Choi, and J Jang, "Web crawler Improvement and Dynamic process Design and Implementation for Effective Data Collection," JKIICE, Vol. 26, No. 11, pp. 1729-1740, 2022.
  12. KOBART, https://github.com/SKT-AI/KoBART.
  13. M. Lewis, Y. Liu, N. Goyal, M. Ghazvininejad, A. Mohamed, O. Levy, V. Stoyanov, and L. Zettlemoyer, "BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension," arXiv:1910.13461, 2019.
  14. Open-source Korean Text Processor, github.com/open-korean-text/open-korean-text.
  15. S. J. Kim, "Exploring the Possibility of Using N-gram Feature for Automatic Scoring of Argumentative Writing Task," Writing Research, No. 41, pp. 37-62, 2019.
  16. H. Shin and J. Choi, "Analysis of User Reviews for Webtoon Applications Using Text Mining," JCCT, Vol. 8, No. 4, pp. 457-468, July 2022.
  17. S. J. Lee and H. J. Kim, "Keyword Extraction from News Corpus using Modified TF-IDF," Journal of Society for e-Business Studies, Vol. 14, No. 4, pp. 59-73, 2009.
  18. H. S. Lee, "Rearch of Late Adolcent Activity based on Using Big Data Analysis," IJACT, Vol. 10, No. 4, pp. 361-368, Dec. 2022.