<![CDATA[Background: Chatbot systems in education, customer service, and social-media platforms increasingly rely on scripted empathy language, yet how these linguistic signs are semiotically constructed and interpreted by users remains empirically under-examined, particularly because prior studies favor computational sentiment metrics over sign-based interpretation. Purpose: This study analyzes how linguistic signs in text-based chatbot responses construct emotional meaning, by identifying the operational function of Saussurean, Peircean, Barthesian, and Ekmanian categories in a corpus of authentic chatbot–user exchanges. Method: This qualitative descriptive study applied semiotic analysis to 120 chatbot–user interaction excerpts purposively collected from three educational chatbots, two customer-service chatbots, and two social-media conversational assistants operating in Indonesian and English between January and March 2026. Excerpts were selected using explicit inclusion criteria (presence of emotional linguistic markers, availability of corresponding user response, and public or consented data source) and coded using a discourse coding sheet, a semiotic interpretation matrix, and an Ekman emotion-category rubric, with inter-coder checking by two independent coders. Results and Discussion: Empathy (21.7%) and reassurance (18.3%) emerged as the most frequent emotional sign categories among the seven coded types, operating as Peircean symbols such as “I understand your concern” and “Thank you for your patience.” These signs were interpreted by participant excerpts as markers of emotional closeness, while emotionally neglectful responses (8.3% of the corpus) were associated with symbolic distance in the surrounding discourse. These findings extend prior sentiment-focused chatbot research by showing specifically how classical semiotic categories operate in AI-generated emotional discourse. Conclusions and Implications: Emotional communication in chatbot interaction is represented through linguistic signs, symbolic interpretation, and contextual discourse rather than through biological affect. The findings contribute to digital semiotics and AI linguistics, and offer a coding framework that developers and researchers can use to design more empathetic and ethically responsible conversational AI.]]>
