Pooja Mithoo1*, Manoj Kumar
In every nation, there are many states which adopt or influence neighboring or far off states with their policy and criminal activities. Many famous news magazines such as ”Washington Monthly”, ”Washington Post”, ”The New York Times ” and ”The Seattle Times” are filled with the activities related to policies drafted by the department of justice, violent crimes, and property crimes. The question is, ”Can we create a model which can deduce rules stating which two or more states influence each other based on the enormous corpus of information?”. The main hurdle in this process, how to assess the impact of a positive entity ”Policy” over the negative entity ”Crime”. We provide a careful procedure of data transformation followed by rule- based deduction and applied FP-Growth for generating the itemsets. We appreciate our itemsets based on the clustering done on our data set using the category of crimes and states as the centroids. We were able to prove that Washington’s policy affects Alaska 89% of the time and crime in Alaska affect Washington 34% of the time.
