A Hierarchical Aggregation Framework for Efficient Multilevel Visual Exploration and Analysis

摘要: Data exploration and visualization systems are of great importance in the Big era, which volume heterogeneity available information make it difficult for humans to manually explore analyse data. Most traditional operate an offline way, limited accessing preprocessed (static) sets They also restrict themselves dealing with small dataset sizes, can be easily handled conventional techniques. However, era has realized availability a amount variety big datasets that dynamic nature; most them offer API or query endpoints online access, data is received stream fashion. Therefore, modern must address challenge on-the-fly scalable visualizations over large data, offering efficient techniques, as well mechanisms abstraction summarization. In this work, we present generic model personalized multilevel analysis numeric temporal Our built on top lightweight tree-based structure efficiently constructed given set This tree aggregates input objects into hierarchical multiscale model. Considering different scenarios datasets, proposed enables exploration, incremental construction prefetching via user interaction, adaptation hierarchies based preferences. A thorough theoretical presented, illustrating efficiency The web-based prototype tool, called SynopsViz offers visual Linked datasets.