The spread of epidemics: social media Analysis VS analysis of Google queries for Flu
Material posted: Publication date: 03-11-2013

Have you heard about Google Flu? By analyzing search queries, Google builds a graph of the trends in "incidence" in time. We decided to test Google data, comparing them with the number of mentions of various respiratory symptoms in social media in Russia: when and where recorded epidemics who are often sick and on what complain? Answers under the cut.

Period: 01.09.13 – 31.10.13 were Analyzed messages in social networks Vkontakte, Twitter, Facebook, LiveJournal, etc. from Russia by means of system of Brand Analytics. Messages with words and phrases "cough", "sore throat", "fever", "body ache" acted as indicators of colds. During this period there were more than 220 thousands of such messages from Russia from 165 thousand authors.

Russian authors during the "normal" period, the average write 2.500 – 3.000 messages a daythat contain mentions of symptoms of colds. The sharp increase in the number of such messages is observed in the period from 7 to 22 September, indicating that the period of seasonal colds. Peaking in September 12 – 7.844 messages.

Abrupt onset of autumn this year played a role, and, predictably, provoked the outbreak.

Comparison with the results of Google Flu

The results of our study, firstly, correlated with data from Google Flu: period maximum number of requests falls on 8-22 September. Secondly, social media data are more correct over time — note the correlation between complaints about health in social networks and search queries:

  • September 12, we fix the peak of messages in social media about disease symptoms,
  • September 15 (i.e. 3 days later) griphouse exhausted trying to find a cure in Google.

To verify our assumption (we piously believe in him, but the prognostics are not a fortune teller and demands of the facts), we will follow the monitoring and promise to share the results of observations and inform in advance when it is better to keep yourself and your family from ordinary flu epidemics and acute respiratory disease.


It is interesting to analyze the peculiarities of the disease on the map of Russia. We investigated the periods of seasonal diseases in different cities of our country and came to two basic conclusions: a) the peak of the disease was observed in all cities, but a different strength, and b) everywhere the period of seasonal diseases comes at the same time. For example, the peak of the disease in Moscow pronounced and falls on 12 September, and the period of seasonal diseases coincides with the period of the disease in Russia.

In St. Petersburg shows an increase in the disease since 7 September, but in comparison with Moscow, this increase is not as dramatic. Although the peak of the disease also can be observed on the 2nd week of September.

And finally, the city with the most pronounced period of seasonal colds – Novosibirsk. The graph shows the increase of cases from 8 to 30 September, the peak of disease (on 16 September), which has the largest power in comparison with Moscow and St. Petersburg.

The most common complaints

This year, in the period of seasonal diseases (September) social media, the authors often complained about the following aspects of your health:

  1. Sore throat – 16,65% of the total number of messages
  2. Fever and fever – 8,52% of the total number of messages
  3. Headache – 2,43% of the total number of messages
  4. Cough – 2,38% of the total number of messages
  5. Aching body – 1,55% of the total number of messages
  6. Chilly – 1,14% of the total number of messages

Be healthy!


Tags: innovation