Professor at Northeastern University, Lu Wang, believes that given a linguistic analysis, artificial intelligence, and data visualization, we can have debates that give us all more meaning to what is going on. To understand what it is that makes that an argument is persuasive is what the academic thinks is the heart of an interdisciplinary project that, in fact, she leads. The ultimate goal is to help the platforms of social networks to evolve from discourses of hatred, of “fake news”, to move on to constructive conversations.
“The discussion should be mechanisms to discover new things about the world,” says Nick Beauchamp, an assistant professor of political science at the university mentioned and a contributor to the project. “the hope is that at the end of a debate the people are not left with only 3l set of new facts that they learned, but get a better way of thinking about the issues raised”.
With this in mind, Wang and Beauchamp designed an algorithm that identifies the characteristics of a strong argument. Using a data set of 118 discussion the Oxford style, where the winner was who could have more audience on your side, the algorithm was able to predict the winner of the debate by 74% of the time.
The model found that the arguments are winners is characterized by certain linguistic characteristics, said Wang, who is an assistant professor in the College of Computing and Information Sciences. For example, the pronouns “us” and “them”, were used more frequently in arguments winning against pronouns like “you” and “me”. The researchers also discovered that the winners of the debate by strategically changing topics to keep the conversation focused on your strongest arguments, said Beauchamp.
For example, suppose that you and I are debating on abolishing or not the death penalty. I could use the argument that there are always errors in the inmates already defined the death penalty, which makes that there are deaths of innocents. But instead of discussing that point (which I know is likely to not be able to win it), focus on the strong point in favor of my argument. Maybe answer to say that the death penalty benefits society by acting as a something that stops the crime.
In fact, the researchers found that explore the topic to something that accommodates more it is that predicts in the long run who can win a debate. This is, however, a first result. Wang plans to apply the algorithm to discussions in the social networks to better understand how they develop.
The assume that people want to engage in constructive debate and not just woman as she entertains herself by insulting the interlocutor, that is to say the least, a perception, optimistic, accepts Beauchamp. But if you have a model of what it means to debate an ideal, then this can bosquejarlo in the way as it should be.
Eventually, Wang wants to work with Twitter and Facebook to develop a tool of automatic evaluation to measure the quality of the discussions. Such a tool could include a visual component that will show the result of the conversations more constructive in where the different sides present strong arguments.
Wang thinks that this research could be put on the public attention to the groups that normally have almost no representation in, and influence, and then you can give voice in the debates of the social networks. “Our tool might be able to help to maintain constructive discussions on where certain groups dominate the conversation while other opinions are ignored,” concludes Wang.
In our country, the political campaigns are all about to end, and unfortunately, we have not distantly a tool such as the Wang and Beauchamp are considering. With it perhaps we would not have to swallow the lies and false arguments of all the parties after the debates, where all claim to have won.