Abstract: Charts effectively convey information through visual logic, providing an intuitive pathway for understanding complex data with obscure relationships and playing a crucial role in scientific documentation. However, the richness of a chart’s visual structure and the complexity of its visual semantics present challenges in creating chart visualizations, while the inherent diversity of underlying data and differences in users’ cognitive perspectives bring divergences in interpreting chart information. With the emergence of automatic chart parsing technologies, an accurate and effective means has been established to interpret charts, laying the groundwork for more advanced insights and reasoning, and further facilitating decision-making. This paper provides a review of research on parsing methods for big data of Chart. It starts by elucidating the necessity and complexity of the chart parsing process from both the perspectives of the chart itself and users. It then outlines the research methods and cutting-edge technologies in chart parsing at three levels: chart perception, structural parsing, and insight reasoning. Furthermore, it elaborates on the downstream applications of chart parsing, including retrieval, interaction, and generation. Finally, the paper summarizes the challenges faced by chart parsing methods, particularly in understanding complex chart types and ensuring information completeness, and presents prospects for future developments in the field.