Abstract. 

The importance of reliable predicting of oil price dynamics is determined by the importance of the oil industry for the Russian economy. This paper proposes a methodology for processing news from different countries, which allows 1) to assess their positive or negative impact on prices; (b) to quantify the importance of oil industry news from different countries. The paper considers news from 23 countries of the world for the period of 2016-2018 with the one month step. 200 descriptors (terms) for each country circulating on the Internet and associated with WTI and Brent oil are selected. The concept of rating and anti-rating of news is introduced, which are calculated on the basis of number of significant terms from the mentioned sets, and degree of their positive or negative correlation with price dynamics. The significance of the introduced concepts is proved by the coincidence of the leading countries in terms of rating and anti-rating of their news with the countries leading on the oil market. The proposed methodology can clarify quantitative forecasts of formal models and provide additional information for experts who work with media.

Keywords: 

mentions in Internet, top sample, news rating, WTI, Brent

PP. 50-56.

DOI: 10.14357/20790279210207

 

References

1. Bushuev V.V., Konoplyanik A.A, Mirkin Ya.M. 2013. Tseny na neft: analiz, tendenciii, prognoz [Oil prices: analysis, tendencies, forecast]. M.: “Energia” Publ. 344 p.

2. Internet resource Yandex: https:// wordstat.yandex.ru/

3. Internet resource Google: https:// trends.google.ru/trends/

4. Boldyreva A. 2016. Postroenie prognoznykh modelei ekonomicheskoi I socialnoi konyunktury po intensivnosti zaprosov v poiskovoi seti Internet [Construction of predictive models for economic and social environment based on intensity of queries to the Internet search machine]. Sovremennaya ekonomika: teoria, politika, innovatsii. Sbornik studencheskikh nauchnykh rabot [Modern economics: theory, politics, innovation. Collection of student research papers]. M.: RANEPA. P. 36–61.

5. Boldyreva A., Sobolevskiy O., Alexandrov M, Danilova V. 2017. Creating collections of descriptors based on Internet querie. Proc. of 15-th Mexican Intern. Conf. on Artificial Intelligence, Springer, LNAI. Vol. 10061, P. 303- 314, Web: // https://link.springer.com/chapt er/10.1007/978-3-319-62434-1_26