Personalized Persuasive Text Generation

Abstract

Persuasion has been defined as a purposeful attempt to change attitudes or behaviors in many organizations that are trying to influence the beliefs, opinions, decisions, or behaviors of their audiences. In various areas, persuasive messages have demonstrated their effectiveness. In order to be persuasive, these messages use psychological strategies. Using strategies can be effective, but their efficacy varies for each individual. For maximum effect, these messages can be tailored to each person based on their personal information and related strategies. During the past few years, natural language text generation methods have improved for automated text generation. However, according to the literature review, no method exists to intelligently generate text for personalized persuasion. In addition, most persuasive technologies are developed with the aim of changing user behavior to achieve user persuasion over time, while changing individual attitudes with just one text has not been sufficiently addressed. By using the CATGAN model and considering the personality of the audience, we present a Personalized Persuasive Text Generation System that achieves state-of-the-art results in a personalized persuasion text generation.