Core Theories Applied in Chatbot Context

Despite the enormous effort to understand the factors driving chatbotChatbot effectiveness, researchers are not univocal. The profound literature audit we performed demonstrated that various theories have been employed in order to identify the key drivers of chatbotChatbot efficiency. Utilitarian (i.e., cognitive related), hedonic (emotion related), and social components emerged to shape the chatbotChatbot performance evaluation, as summarised in the thematic map and taxonomy we developed (see Chap. 2 ). The core theoretical notions are organised around three main pillars: acceptanceAcceptance models, behavioural theories, and social influenceSocial influence theories. The acceptanceAcceptance models included are: TAMTechnology Acceptance Model (TAM), UTAU, Diffusion of InnovationDiffusion of Innovation (DOI), Gratification theoryGratification theory, and Uncanny Valley theoryUncanny valley theory. Frameworks like the Planned behaviour, Reasoned action, Self-determination, Motivation, and Big five are among the most frequently cited in the papers audited and thus are discussed hereby in detail. In addition to these theories, anthropomorphismAnthropomorphism, social agency, social presenceSocial presence, social response, parasocial interaction, and CASAComputers as Social Actors (CASA) theories are brought to the table. Fundamental paradigms, with relevant examples and related papers, are summarised in taxonomy of core theories applied in chatbotChatbot context (see Fig. 3.1) and presented in detail below.