Voting algorithm in the play Julius Caesar

摘要

This paper suggests a voting algorithm for predicting people's choices. Usually, once a new algorithm is offered, one needs to prove the soundness of the algorithm, i.e., showing that the algorithm does the thing it is set up to do. But in the case of election prediction algorithms it's not clear how to prove their soundness. This paper offers a way to deal with this problem by analysing the social networks of plays, following Shakespeare dictum: 'all the world is a stage, and all the men and women merely players', As you like it, act I scene VII. The advantage of this approach is clear. The story of Julius Caesar is part of our collective memory. We learn the leading characters views and opinions throughout the play. For example, we know that Brutus was against Caesar, while Antony supported Caesar, and opposed Brutus. By generating the social graph from a play, the problems of soundness is solved, which makes the process more scientific. The voting algorithm needs anchors for each party. The anchors are the charactersodes whose opinions are clear and known. In the case of the play Julius Caesar, the anchors are Brutus and Cassius for the conspirators and Antony and Octavius for the supporters of Caesar. After the anchors were identified the voting algorithm uses simple random walks, to divide the network of characters into two parties. Now it is possible to examine the output of the voting algorithm and see that indeed the algorithm works correctly.