I was keen to write this blog was after attending Ting Li’s lecture on how Twitter has an impact the stock market of a company. The fact that a major company like The Associated Press faced a huge loss in a matter of four minutes due to its twitter account being hacked is truly amusing!
According to a new study, social media sites such as twitter could be used to predict major, potentially world-shaking events. Nathan Kallus, a PhD from Massachusetts Institute of Technology, believed that crowd behavior could be predicted by mining tweets for any future events and then analyzing the trends associated with them. He reviewed the tweets associated with the 2013 social unrest in Egypt and claimed the protests were clearly visible in the Twitter feed days in advance. He added that the social media content even predicted that the protests would go on for weeks beyond the anniversary. He came up with the rationale behind this through natural language processing, which helps in extracting event information from content. This includes; type of event, what entities are involved and in what role, sentiment and tone, and the occurrence time range of the event discussed. Another predictive indicator included the statements mentioned on Twitter about a future date from the time of posting.
The idea of twitter predicting the future has been controversial. Kallus made this prediction about a historical event using historical data. However, do you think its possible to pick out these kinds of evidences before the event actually occurred? It could also be very time-consuming and difficult to accurately analyse the event from various social media platforms. There is also the question of reliability of the information posted in social media. Social media trends tend to vary and might lead to incorrect data. Additionally, I don’t believe that tweets truly represent the intentions of the entire population. Thus, I personally don’t think social media could help in accurately predicting major events.
What is your opinion in this matter? Do you think that sometimes social media could be useful in predicting certain events like box office revenues for movies?
http://www.fastcompany.com/3026593/fast-feed/can-twitter-predict-major-events-a-new-study-says-yes
http://www.technologyreview.com/view/524871/can-twitter-predict-major-events-such-as-mass-protests/
http://www.technologyreview.com/view/427807/twitter-cannot-predict-elections-either/
It is a very interesting topic to have a look at because of its predictive value. I believe it is hard to actually predict certain events that would normally not take place, like the example of social unrest in Egypt. This is hard to predict because people were probably not aware of the social media activity around it before it happened. Afterwards people can tell that it could have been predicted, but if you want to predict it beforehand, you have to keep up with every single thing that could take place or happen in the world. That is just very time consuming and probably you are searching for something that ain’t happening.
The example mentioned at the end about box office revenues for movies is easier to predict. Each movie has a release date that people know about and therefore people know it is coming up. The previews also create a certain hype around it before the movie is actually there. I must say that the amount of likes or tweets does not necessarily mean someone is going to watch the movie, but the more people tweet about a movie in a positive way, I think we could say more people would probably go there.
So if you ask me, I would say that the predictive value of Twitter is mainly there for expected occasions, of which people are sure they are happening and know when they will happen.