Northeastern University Background
Companies, organizations, and political groups spend millions of dollars on persuasive campaigns. Field experts perform research to identify the potential arguments, for or against their position, who believes those arguments and which one’s people find most persuasive. This is a labor-intensive process which includes reading expert articles, conversations online (i.e., Twitter and Reddit threads).
Researchers at Northeastern have developed an optimized method for identifying persuasive arguments. Rather than relying on an expensive, labor-intensive process to identify persuasive arguments, this new method pulls out large quantities of online text, identifies argumentative claims within the text and categorizes those arguments via topic modeling. An optimization algorithm then identifies the most persuasive argument, leveraging workers from Mechanical Turk. This process allows for a more extensive search of arguments, and for the efficient identification of the most persuasive argument.
Enables searches that are:
Faster and more efficient
More effective with greater reach
Idea validation through customer discovery
Testing and validating MVP with pilot partner(s)