User:Bwlynch
Real-World Predictions with Social Media
[edit]![](http://upload.wikimedia.org/wikipedia/commons/thumb/e/e6/Twitterbird.png/220px-Twitterbird.png)
As Twitter grows and expands, so too does the data that it brings. The wide array of tweets, replies, favorites, and more pave the way for Twitter data to have a wide array of applications. One application that has gotten increased attention in recent times is predicting elections with Twitter. Everyone is looking for a way to predict who will win the election, and one can look to a group of German researchers who used Twitter to see if elections could be predicted with it, and found strong data suggesting so.[1] Similarly to the predicting of elections, Twitter can also be used to track public opinion; Twitter data reveals that Twitter provides many useful insights in various areas, and could be used in conjunction with traditional polling.[2] Social can also be used in areas outside of the scope of public opinion. A group of researchers found that blog posts can be correlated to stock markets, and could even semi-reliably predict in which direction stocks would move. [3] Twitter also has uses in the medical field. Twitter data was used to track those with self-reported mental health issues, and it provided a window into daily life for these patients. [4] Twitter data revealed how seasonal depression could be tracked across a variety of cultures using features contained within the tweets too [5] In addition, Twitter has been used for more nefarious purposes, but Twitter data can be used to identify those who are lying on Twitter for a variety of reasons. When Superstorm Sandy hit New York, some users of Twitter attempted to fake their location to make it appear as if they were there. However, researchers were able to use Twitter data to weed out these Twitter fakes, delving deep into Twitter’s API to locate these users [6]
Twitter and Predicting Elections
[edit]One area that Twitter is seeing increased research in is in the study of elections. Specifically, Twitter was theorized to be able to be used as a place for effective debate, representative of the political atmosphere, and able to predict elections. Four German researchers analyzed tweets from an election-time period in Germany, and saw how they could be applied to three political areas. The data revealed that all three initial areas of theory were proven correct, as the tweets were able to support the claims that all three statements were accurate based on the singular election.[7]
Twitter and Popular Opinions
[edit]Twitter has also been theorized to have use in determining public opinion, perhaps even being used in conjunction with traditional polling. So far, the four areas of general unsolicited opinions, job approval rating for President Obama, consumer satisfaction rating, and business sentiment shifts have been identified as able to be determined on Twitter. Researchers analyzed the "happiness" of a variety of keywords relating to these subjects, and compared Twitter data to actual polling data when necessary. In the future, Twitter may be proven to be useful in a wide variety of areas in addition to these.[8]
![](http://upload.wikimedia.org/wikipedia/commons/thumb/6/69/Stock_Market_Changes.png/220px-Stock_Market_Changes.png)
Stock Markets and Blogs
[edit]Another area which has seen increased research is social media and the stock markets. One group of researchers looked at the stock market in the context of blog posts and comments, the first of its kind. Researchers gathered data from the stock markets and blogs, and created multiple tests to run the blog data through. While two of the baseline tests failed to show any correlation, more advanced regression techniques did show some correlation. These techniques could predict the magnitude of stock shift relatively accurately, and they were quite accurate in predicting the direction of stock shifts.[9]
- ^ Tumasjan, Andranik, et al. "Predicting Elections with Twitter: What 140 Characters Reveal about Political Sentiment." ICWSM 10 (2010): 178-185..
- ^ Cody, Emily M., et al. "Public Opinion Polling with Twitter." arXiv preprintarXiv:1608.02024 (2016)..
- ^ Ajita, Munmun De Choudhury Hari Sundaram, and John Dorée Duncan Seligmann. "Can Blog Communication Dynamics be correlated with Stock Market Activity?.".
- ^ Coppersmith, Glen, et al. "From ADHD to SAD: Analyzing the language of mental health on Twitter through self-reported diagnoses." Proceedings of the 2nd Workshop on Computational Linguistics and Clinical Psychology: From Linguistic Signal to Clinical Reality. 2015..
- ^ Golder, Scott A., and Michael W. Macy. "Diurnal and seasonal mood vary with work, sleep, and daylength across diverse cultures." Science 333.6051 (2011): 1878-1881..
- ^ Aditi, et al. "Faking sandy: characterizing and identifying fake images on twitter during hurricane sandy." Proceedings of the 22nd international conference on World Wide Web. ACM, 2013..
- ^ Tumasjan, Andranik, et al. "Predicting Elections with Twitter: What 140 Characters Reveal about Political Sentiment." ICWSM 10 (2010): 178-185..
- ^ Cody, Emily M., et al. "Public Opinion Polling with Twitter." arXiv preprintarXiv:1608.02024 (2016)..
- ^ Ajita, Munmun De Choudhury Hari Sundaram, and John Dorée Duncan Seligmann. "Can Blog Communication Dynamics be correlated with Stock Market Activity?.".