/pdf/preconceito-linguistico-o-que-e-como-se-faz-2dgkn7pd9n.pdf

Preconceito lingüístico: o que é, como se faz

Evaluation of Human–Computer Interaction (HCI) in a textual way is a recent type of evaluation, demanding studies of techniques to improve it. This work aims to investigate the automatic classification of the polarity of opinion in postings related to the use of the system. We analyzed 1,345 postings of an academic system with social characteristics and performed two investigations: one uses an automatic classifier and the other applies Data Mining algorithms. As results, we present the most relevant words of each polarity, their characteristics and a discussion about the investigations carried out.

/pdf/investigating-the-polarity-of-user-postings-in-a-social-s6ox4fc6rx.pdf

Investigating the Polarity of User Postings in a Social System

Ethics is a field of study that analyses practices, their principles, customs, and habits. The Grand Research Challenges in Human-Computer Interaction in Brazil (GranDIHC-BR) between 2012 and 2022 highlighted the relevance of the topic Ethics in the context of Human Values. Aligned with that, we claim that computational techniques are not neutral and, therefore, ethical analysis is essential to the design of applications as well as theoretical or conceptual research. The goal of this work is to provide an understanding on how researchers have addressed the ethical aspects by analyzing the publications of the IHC-BR: the Brazilian Symposium on Human Factors in Computing Systems, the leading conference dedicated to HCI in Brazil. Thus, we conducted a Systematic Literature Review guided by the main research question: how does ethics permeate research published between 2006 and 2021? As a result, although less than 5% of the papers discussed ethical content, we observed a significant growth in ethical aspects over the years, both associated with the meta-research and its application.

/pdf/a-meta-scientific-broad-panorama-of-ethical-aspects-in-the-28qeyudl.pdf

A meta-scientific broad panorama of ethical aspects in the Brazilian IHC

Ubiquitous systems are made up of sensors and smart devices that communicate with each other to achieve certain goals. They have specific characteristics, such as transparency, attention, calmness, mobility, and context-awareness. Given that the Human-Computer Interaction (HCI) evaluation methods have been proposed for traditional systems, these characteristics present challenges that may require their adaptation. In this paper, we conducted an exploratory study aimed to compare the suitability of Heuristic Evaluation and the MALTU Model to evaluate the Usability and User eXperience (UX) of ubiquitous systems. For this, we apply both methods for evaluating a healthcare system and perform a comparative analysis. From our findings, MALTU is indicated for UX evaluation, being a low-cost approach when compared to Heuristic Evaluation, which is indicated for usability evaluation. Both methods were able to identify the problems that emerge from ubiquity and can provide better results when applied in aggregate. Furthermore, we provide recommendations to apply both methods for the Usability and UX evaluation of ubiquitous systems.

Comparing heuristic evaluation and MALTU model in interaction evaluation of ubiquitous systems

Human Values, the fourth challenge listed in the Grand Research Challenges in Human-Computer Interaction in Brazil (GranDIHC-BR), between 2012 and 2022, presents aspects of Ethics alongside Privacy and Post-death Digital Legacy. The field of Ethics analyses practices, their principles, customs, and habits. Since computational techniques are not neutral and, to a large extent, they are embargoed from technological determinisms, ethical analysis and evaluation are necessary to design computational solutions and theoretical or conceptual research. Eight years after the publication of the GranDIHC-BR, this work aims to understand how researchers have addressed the ethical aspects by analyzing the publications of the IHC-BR: the Brazilian Symposium on Human Factors in Computing Systems, the leading conference dedicated to HCI in Brazil. Thus, we conducted a Systematic Literature Review guided by the main research question: how does ethics permeate research published between 2006 and 2020? As a result, although less than 5% of the papers discussed ethical content, we observed a significant growth in ethical aspects over the years, both associated with the epistemology of research and its application.

Ethics: What is the Research Scenario in the Brazilian Symposium IHC?

The choice of apps by users in App Stores is influenced by other users' ratings (reviews and star ratings). However, it is unclear what the relationship between user appraisal and an Usability and User eXperience (UUX) evaluation. In this paper, we will make the following investigations: i) what is the result of an UUX evaluation of mobile applications using user reviews in App Stores? ii) does the star rating represent the opinions of users about an app? and iii) is there a relationship between the ranking of apps and an evaluation of the UUX? Some results: i) the result of an UUX evaluation was different for the same app on different systems; ii) the star rating was in agreement with the opinions of the users, in two of the three apps tested; and iii) the results of an UUX evaluation are similar to the ranking of apps.

Beyond traditional evaluations: user's view in app stores

Assistentes de Voz Inteligentes (AVI) estao se tornando cada vez mais populares em todo o mundo, provendo facil interacao para realizacao de servicos e acessos a dispositivos. No entanto, o treinamento desses assistentes de voz pode estar limitado a falas padronizadas gerando o risco de desestimular a adocao desta tecnologia inclusiva por grupos com caracteristicas de fala distintas da padrao. A eventual diferenca de qualidade de desempenho dos assistentes em virtude das formas especificas da fala de diferentes grupos de pessoas nos paises em desenvolvimento pode ser considerado um vies, bem como um fator que discriminara principalmente os grupos mais vulneraveis e, consequentemente, aumentara a lacuna digital nestes paises. Neste trabalho, apresentamos um estudo de caso que analisa a presenca de vies na interacao por voz com os assistentes, ou seja, uma diferenca na qualidade do desempenho dadas caracteristicas especificas de fala dos usuarios. Avaliamos o comportamento do Google Assistente e da Siri para grupos de pessoas distribuidas por genero e regiao do pais, avaliando duas caracteristicas da linguagem: a cacoepia e a disfluencia. Resultados preliminares indicam que a cacoepia e a disfluencia estao relacionadas com a regiao de origem do falante e tem impacto no desempenho dos assistentes de voz.

Discrimination analysis of intelligent voice assistants

This paper presents a textual evaluation experience in Usability and User eXperience of Spotify application. The methodology used for the textual evaluation was MALTU. In this study, we collected 100 users' postings from a popular website (Reclame Aqui) and 7 participants classified postings and discussed their impressions about this form of evaluation.

Evaluation of Spotify: an evaluation textual experience using the Maltu methodology

One of the ways to evaluate a system is from Textual Evaluation. This type of evaluation consider the textual user’s opinions to infer aspects of the interaction with the system. Although this method covers many texts and consi- ders spontaneous narratives of the users, it takes a lot of time and effort. Some authors have reported on the need to compare evaluations techniques in order to investigate their effectiveness in revealing issue or to supplement the results of a systems assessment. This study presents a comparative analysis between the textual evaluation and user testing. A case study was performed evaluating the usability and user experience of a health app. As a result, the techniques were analyzed based on aspects that involved describing the results, resources needed and description of problems and users.

/pdf/textual-evaluation-vs-user-testing-a-comparative-analysis-4133plc1e7.pdf

Textual evaluation vs. User testing: a comparative analysis

A tecnologia tem apoiado professores em diversas areas. O projeto
de softwares educativos envolve o conhecimento do contexto de uso em que
ele sera inserido. No entanto, nem sempre as tecnicas de projeto de softwares
focam no envolvimento de seus usuarios-alvo. Este artigo apresenta o projeto
de uma aplicacao para apoiar professores no ensino de ingles. O projeto foi
desenvolvido utilizando uma metodologia de design participativo que sugere a
experiencia de uso como forma de coleta e exploracao de ideias para projetar
um sistema. Um prototipo de alta fidelidade foi produzido e avaliado. Como
resultados, apresenta-se o processo de construcao do aplicativo desde as ideias
iniciais ate as percepcoes finais dos usuarios, depois do aplicativo estar pronto.

/pdf/projetando-uma-aplicacao-para-apoiar-o-ensino-de-ingles-a-48uz9ht4dr.pdf

Thiago Hellen O. da Silva

Papers

Beyond traditional evaluations: user's view in app stores

Discrimination analysis of intelligent voice assistants

Evaluation of Spotify: an evaluation textual experience using the Maltu methodology

Textual evaluation vs. User testing: a comparative analysis

Projetando uma aplicação para apoiar o ensino de inglês a partir de uma experiência de uso