A Systematic Review of Symbolic Aggregate Approximation (SAX)

Abstract

Time series data can be analyzed through various techniques to tackle classification or regression tasks. Symbolic Aggregate Approximation (SAX) is one such technique used for time series data reduction that converts the data into a symbolic representation, enabling more efficient storage, retrieval, and analysis by reducing the dimensionality while preserving the essential patterns within the time series. In this paper, we provide a systematic literature review of SAX by examining relevant literature from 2007 to 2025. The review includes 321 articles sourced from the Web of Science (WOS) database. However, the 85 most cited and recently published studies are summarized. Utilizing collaboration network analysis, the study identifies the nations, affiliations, and authors involved in SAX research, as well as their co-authors and commonalities. Additionally, an analysis is conducted to explore the potential relationship between the articles and the United Nations' Sustainable Development Goals. These findings provide insights into the current landscape of SAX research and offer potential avenues for future exploration. By pinpointing research gaps, scholars can use this review to anticipate forthcoming research trajectories.