Avaliação da qualidade de dados ambientais por meio de técnicas de analítica visual

Abstract

The technological advances has boosted automated and periodic data collections using sensors, such as those performed in Integrated Environmental Monitoring System (SIMA), which dataset is used in this work. Despite of the automation degree, there are some flaws in the dataset caused by problems in the collection, transmission or data storage. Due to the existence of large multivariate temporal data and the possibility of failure, interactive visual representations of the data can improve the study and characterization of the dataset. In this work, Visual Analytics techniques were applied to obtain some characteristics from the environmental dataset, which subsequently can impact the quality of analysis of the associated phenomena. The results showed the benefits of using visual and interactive representations for dataset analysis, which facilitates the information extraction about dataset characteristics, such as data availability, sensors operation characteristics, and evidences of a failure pattern.