OCEANS-6 : Character Trait, Age Demographic Analysis and Sexist Speech Detection of Dialogues

Character trait analysis has always been an open-ended problem with many potential applications. In a technologically advanced age where every industry seeks to benefit from artificial intelligence, combining its capacity with NLP offers us a fresh way to gather data on people. The majority of features in the digital world are typically observed based on the user’s interests and other incredibly hard-to-gather information. The authors examine a novel unsupervised graph-based method that can identify useful information about a person’s speech or writing style. The study encompasses three major components: Character Trait Prediction, Age Demographic Analysis, and Sexist Speech Detection. Every person is categorised into one of five character traits (OCEAN - Big Five character traits, which include openness, conscientiousness, extroversion, agreeableness, and neuroticism), with the addition of a sixth dimension: a non-binary sexist speech, making it the OCEANS-6. The proposed algorithm removes the need for expertly labelled data and offers new angles on the character analysis issue. This work offers a multifaceted approach to dialogue analysis, with applications ranging from character trait prediction in literature to age demographic analysis and the detection of sexist speech in social media. The findings contribute to a deeper understanding of character dynamics, linguistic cues, and social discourse in various contexts.