Análise temporal de postagens relacionadas ao uso do sistema de gerenciamento acadêmico SIGAA

Abstract

The emergence of Social Systems provided several ways of evaluating systems. The textual evaluation, for example, considers the process of selection and analysis public posts aiming at improving usability and user experience. These posts, in which the user refers to the system in use are called PRUs (Posts Related to Use). The textual evaluation, as well as the other evaluations, present the results of the moment they were performed, and the difficulties encountered at the beginning of the interaction can be minimized or maximized over time. Therefore, to investigate whether user satisfaction increases or decreases, or if doubts become more frequent, it is necessary to conduct more than one evaluation in order to compare the results. This work presents an algorithm for temporal analysis of PRU aspects, such as polarity (positive and negative) and types of PRUs (praise, criticism, doubt, comparison and suggestion). For this, an experiment was carried out with 6,600 postings of an academic management system, from 2010 until 2011 and written in Portuguese of Brazil. After the collection, a classification was performed to obtain the PRUs and classify them in the following categories suggested by MALTU (Methodology for Evaluation of interaction in social systems from the User’s Textual Language): a) polarity; b) functionality; and c) type of PRU. These postings were analyzed by the algorithm and the following results were obtained: 1) quantitative postings over the years and months; 2) quantitative of each category over the years and months; 3) comparing the quantitative attributes of each category over the years and months and 4) quantifying the