Chatbots: um estudo sobre sistemas conversacionais e a experiência das pessoas usuárias sob a perspectiva de uma nova interação

Abstract

The increasing use of smartphones has encouraged the development of new systems capable of meeting different demands more conveniently - mobile applications. These new software solutions have driven and enabled new economic models in a short period of time, but they have also begun to occupy mobile phone memory space. Given this scenario and the growing adoption of instant text messaging applications, conversational systems technology (chatbots) has experienced exponential growth over the past decade. Therefore, this research is a case study with a qualitative approach, focusing on the chatbot of a telecommunications company, as preliminary studies have indicated that customer service chatbots have boosted the retail sector, especially in the telecommunications segment. The main objective of the study was to investigate the perceptions that conversational automations, such as chatbots, provide to users in carrying out their tasks. The research was divided into 4 stages: In the first stage, named 'review,' criteria for task success and completeness were understood based on the literature through a systematic literature review. The next stage, 'investigation,' involved applying a survey questionnaire to examine people's perceptions in interacting with chatbots, testing hypotheses to quantitatively assess people's ability to identify a chatbot, the feelings generated by these systems, people's satisfaction and frustration levels, and their tolerance towards limitations and errors made by chatbots. The third stage, named 'experimentation,' evaluated the 'internet connection problems' stage of the chatbot's conversational flow of a telecom operator through heuristic analysis, usability tests, and a user experience questionnaire (UEQ). Eight individuals of different ages and technological maturity participated in the test, performing interaction tasks based on heuristic problems identified by experts. The final stage, 'data analysis,' integrates the findings from the previous 3 stages. Based on the collected information, the study proposes experience criteria - parameters to compare and decide the interaction model with the system studied, considering literature guidelines, structural characteristics of chatbots, and feedback from users who participated in usability tests.