Text Mining
With all the debates that the presidential candidates do before the election, usually, it ends up that there is just too much data for any one person to process and analyze. In order for our group, Contra, to attempt to answer the correlation with the War on Drugs and Presidential Elections, we would seek to use data mining software to weed out information that is not pertinent to our project when we analyze presidential debates.
As the War on Drugs is not a widely talked about topic for presidential elections, the use of data mining softwares will be beneficial in helping us find out exactly how often phrases such as “drugs”, “cartels”, “anti-(fill in whatever drug)” come up during the debates. Then we would use this information to come to a conclusion on how much emphasis is put on the War on Drugs during the presidential elections and hopefully we can answer some of the questions that our group posed for our project.
Your idea of searching for phrases related to the WOD is a good start, but I wonder whether the resulting portait will be entirely accurate? Is there is a way the text mining could be designed in a way that privileges certain sources over others? Or would the mention of “drugs” in all sources be treated equally? Is it safe to assume that increased frequency of a phrase such as the “War on Drugs” represents an actual increase in attention being paid to it by the candidates and voters?