S-Paths

Research, Design & Development

2018 - 2020

Marie designed and developed S-Paths, a visualisation tool producing path-based synthetic overviews to browse collections in Knowledge Graphs. A matching algorithm compares the path outlines of the collection with the requirements of a set of views optimised to provide synthetic visualisations at various scales, and provides a default view to users. Starting with an overview of a whole collection, it offers new synthetic views of the subsets as they refine their selection.

Process for generating a default view: once semantic paths for a given resource set have been retrieved, the system evaluates the suitability of the different views to generate a default representation of this resource set.
The default view on Laureates shows their date of birth and their gender on a stacked chart. Switching dimensions to award year and discipline renders a different stacked chart. Switching to a map view offers, as a first choice, the laureates’ birthplace latitude and longitude as the dimensions used to plot Nobel laureates on the map.
Key views of the Nobel database: a 2D density plot showing the count of awards per year (binned by decade) and category; another density plot showing award shares instead of years; a histogram showing the repartition of awards by gender over the years; a map showing the birthplace of award laureates. Now considering laureates as the resource set: a histogram showing laureates’ gender and birth year; an info card detailing all semantic paths for one laureate in the set; and a histogram showing ge
Animated transition from a sub-selection made in a 2D density plot show counts for types of Award File aggregated by decade, to a histogram showing the distribution of prizes per discipline for each individual year.
Brushing & Linking between two views: items selected on the map are highlighted in the histogram.