Three-way data analytics: Preparing and analyzing data in threes

Abstract

Three-way decision and its underlying philosophy of thinking in threes have been investigated in a variety of topics. By observing that many three-way topics are related to data analytics, this work presents three-way data analytics, examine a number of its existing topics, and propose a few potential new topics. Based on the two stages of data analytics (i.e., data preparation and data analysis) and the Data-Information-Knowledge-Wisdom (DIKW) hierarchy, we organize the existing three-way topics into three bunches, namely, three-way data preparation concerning data, three-way data analysis concerning information, and three-way data analysis concerning knowledge and wisdom. This systematic organization may shed light on a new view of understanding current results. Moreover, it may promote the studies on the relationships between topics from different levels and inspire their synthesis to obtain three-way approaches to the complete picture of data analytics. Furthermore, we may discover topics that are commonly studied in conventional data analytics but not with three-way approaches. Since three-way approaches are theoretically and practically proved to be efficient and effective in many existing topics, many new topics may be worth investigating. This work presents a few potential new topics, including three-way data selection, three-way word analysis, and three-way data visualization.