Using and Exploring Hierarchical Data in Spreadsheets

Abstract

More and more data nowadays exist in hierarchical formats such as JSON due to the increasing popularity of web applications and web services. While many end-user systems support getting hierarchical data from databases without programming, they provide very little support for using hierarchical data beyond turning the data into a flat string or table. In this paper, we present a spreadsheet tool for using and exploring hierarchical datasets. We introduce novel interaction techniques and algorithms to manipulate and visualize hierarchical data in a spreadsheet using the data's relative hierarchical relationships with the data in its adjacent columns. Our tool leverages the data's structural information to support selecting, grouping, joining, sorting and filtering hierarchical data in spreadsheets. Our lab study showed that our tool helped spreadsheet users complete data exploration tasks nearly two times faster than using Excel and even outperform programmers in most tasks.

The usability of task modeling tools

“…This suggests that duplicating hierarchies or allowing for the creation of templates would be beneficial in order to populate task models. Bulk editing could also be facilitated if the model was specified using spreadsheets and their typical functionalities including duplication, making minor modifications of entities effectively and arranging the elements of the model hierarchically [13]. Spreadsheets could then be loaded to populate the graphical model for later refinement and model checking purposes.…”

Interactive browsing and navigation in relational databases

“…Although these visualization systems provide an overview of data sets, they are not suited for examining database instances along with attributes. In exploring and analyzing instance-level information, tabular interfaces, including spreadsheets, are better suited and often preferred by database users [24,46,35,17,25]. Tyszkiewicz [46] argued that spreadsheets can play a role as a database engine by using functions and macros.…”

Mavo

“…Quilt [5] extended HTML with a language for binding an arbitrary web page to a Google spreadsheet "back-end", enabling web authors to gain access to lightweight computation without programming. Gneiss [10,11] was a web application within which authors could manage and compute over hierarchical data using an extended spreadsheet metaphor, then use a graphical front end to interact with that data.…”