Posts Tagged: contest

Does an epidemics affects us on a psychological level?

On 19-11-2014 the BigData Community from Cluj-Napoca held the  4th BigData/DataScience Cluj-Napoca meetup and 1st Contest with prizes offered by TrustYou. The contest’s theme was : “Take  GDELT datasets and do crazy things with them!”. This contest gave me the chance to have my first encounter with BigData and Data Analysis [ iei 🙂 ].

 

10625063_10152917940029271_253315311961567841_n

 

Choosing the theme

Observing that the illness or potential risk of our acquaintances sadden us or make us be more aware of that problem, I wanted to see how this is manifested on a larger scale.

Also, a target for this study was to see if in the areas that risk an epidemics psychological aid is needed.

The criteria to choose the country for my study were:

  • close enough to the place Ebola appeared – Papa New Guinea,  in order to be at a potential risk,
  • far enough not to be a 100% target.
  • affected by Ebola in 2014
  • declared Ebola free.

The above criteria made me pick Nigeria for the study.

The study

First question I asked myself was how the psychological impact  could be determined. I took for the impact measurement the  number of suicides.

Using GDELT Analysis Service I simulated a timed heat  map for every week from 30.12.2013 to 20.10.2014 for GDELT GKG Nigeria and suicidal mentions.

The heat map is a graphic representation with the values represented as colours. For GDELT as the number of mentions for the searched  words increases, the colour gets warmer (few mentions are represented as blue, while a large number of mentions is represented by red)

A preview for the study  can be seen in the next gif.

multiple

It can be seen that at the beginning of the year there have not been many suicidal mentions in the media, for example the firth two weeks of February have no such mentions. (Maybe everyone was happy before Valentine’s day 😀 ). After 23th March, when WHO (Word Health Organization) declares the outbreak of Ebola the number of suicidal mentions in Nigeria increases in a visible way.

On 20th July Ebola is introduces in Nigeria. The person who introduced Ebola in Nigeria negated any connections with Ebola, that is the reason he was treated for Meningitis. A week after, he dies from Ebola as well as his taxi driver. Part of his medical stuff gets infected with Ebola and for some the disease is fatal. During the summer further Ebola cases appear. The last Ebola case is resolved on 14th September and 42 days later Nigeria is declared Ebola free.

At the end of the presentation there is the tone timeline for Nigeria in the studied period. The tone timeline graphically represents from extremely positive to extremely negative how emotions, in regards to one subject, change in time. It’s easily seen that when the first deaths of Ebola appear and Ebola is confirmed in Niger the articles that involve Nigeria tend to be more negative.

 

Full presentation can be seen here

Conclusions

Is there the certainty that all these mentions of suicides in Nigeria is linked to Ebola?

No. Other events could have influenced these mentions. This study was just a correlation created between the two events.