Case 7 — Stereotypes
Case 7 — Stereotypes
Doctoral research by Carmela Sportelli — Brillamenti
Model: Rule-based conversational agent + Intervention
Interactive storytelling ➔ Theme: Sexist stereotypes & hate speech
Brillamenti is a web-app built as a conversational story rather than a lecture. Users chat with three characters: Stella, a counter-stereotypical female peer model; Zorp, an alien with a purely logical outside view of human bias; and Zaphod, who voices the naive, sexist positions absorbed from online rhetoric. Guided by Stella, users work to counter Zaphod's distorted beliefs about gender — learning to recognize biased narratives and practising real counterspeech skills, aimed especially at teenagers and young adults.
"A rule-based system is transparent and predictable by design — but that same rigidity limits how far the conversation can flex to a real user. The bet is that a tightly scripted dialogue can still feel safe and spontaneous enough to shift a belief, not just click through a quiz. "
Carmela Sportelli, Uninettuno
Supervisors: Alessandro Pollini (Uninettuno), Marinella Paciello (Sapienza), Francesca D'Errico (Università di Bari)