MALTU – um modelo para avaliação da interação em sistemas sociais a partir da linguagem textual do usuário

• Ano de defesa: 2015
• Autor(a) principal: Mendes, Marília Soares
• Orientador(a): Castro, Miguel Franklin de
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Palavras-chave em Português: Interação humano-computador; Usabilidade; Experiência do usuário; Processamento da linguagem natural; Mineração de dados; Human Computer Interaction (HCI); Usability; User experience
• Link de acesso: http://www.repositorio.ufc.br/handle/riufc/19820

The field of Human Computer Interaction (HCI) has suggested various methods for evaluating systems in order to improve their usability and User eXperience (UX). The advent of Web 2.0 has allowed the development of applications marked by collaboration, communication and interaction among their users in a way and on a scale never seen before. Social Systems (SS) (e.g. Twitter, Facebook, MySpace, LinkedIn etc.) are examples of such applications and have features such as: frequent exchange of messages, spontaneity and expression of feelings. The opportunities and challenges posed by these types of applications require the traditional evaluation methods to be reassessed, taking into consideration these new characteristics. For instance, the postings of users on SS reveal their opinions on various issues, including on what they think of the system. This work aims to test the hypothesis that the postings of users in SS provide relevant data for evaluation of the usability and of UX in SS. While researching through literature, we have not identified any evaluation model intending to collect and interpret texts from users in order to assess the user experience and system usability. Thus, this thesis proposes MALTU - Model for evaluation of interaction in social systems from the Users Textual Language. In order to provide a basis for the development of the proposed model, we conducted a study of how users express their opinions on the system in natural language. We extracted postings of users from four SS of different contexts. HCI experts classified, studied and processed such postings by using Natural Language Processing (PLN) techniques and data mining, and then analyzed them in order to obtain a generic model. The MALTU was applied in two SS: an entertainment and an educational SS. The results show that is possible to evaluate a system from the postings of users in SS. Such assessments are aided by extraction patterns related to the use, to the types of postings and to HCI factors used in system.