8.10.2021 | Akademie Mainz
ISS 2021
Graph Tools
for data exploration and indexing
Slides: https://studiengang-digitale-methodik.pages.gitlab.rlp.net/modul-5/5c/slides/graphtools
Julian Jarosch |
@digicademy |
digicademy |
CC-BY 4.0
Table of Contents
- Requirements for graph data exploration tools
- Exploration of aggregated data in SemSpect
- Overview of available tools
- Evaluation of graph data exploration tools
01
Requirements for graph data exploration tools
some requirements
for graph data exploration tools
general aims:
- efficient overviews
- concurrent exploration of multiple result sets / browsing paths
- provoke information discovery
technical requirements:
- interactive refinement of results
- clustering of large result sets
- approximate queries / queries from examples
(partly excerpted from Mottin, Jentzsch & Müller (2016))
02
Exploration of aggregated data in SemSpect
SemSpect
software
Web app with a focus on aggregating large property graphs, so that their dominant structures emerge
- licensing
- proprietary, commercial
- free access for scientific use may be requested – get in touch directly
-
technology
- back end: neo4j
- front end: HTML5, JavaScript
03
Overview of available tools
available tools
survey of current state-of-the-art
Explore the tools and their capabilities in groups. Take note
- if / how the requirements formulated previously apply
- what advantages / disadvantages you see
in the respective tool.
-
SemSpect
(best use the GovTrack demo, some of the others have errors) -
Wikidata
-
Wikidata Query Service
(e.g. example query “groups of characters in the Marvel universe”) -
Wikidata Graph Builder
(e.g. example graph “children of Agnes of the Palatinate”) -
Wikidata Lexeme Graph Builder
(e.g. example graph words derived from Proto-Indo-European “*wódr̥”)
-
Wikidata Query Service
We won’t go into these today, you can look at them on your own time:
- graphVizdb (e.g. dataset “DBpedia Person”)
- inWalk (e.g. dataset “Twitter News”)
04
Evaluation of graph data exploration tools
some requirements
for graph data exploration tools
general aims:
- efficient overviews
- concurrent exploration of multiple result sets / browsing paths
- provokes information discovery
technical requirements:
- interactive refinement of results
- clustering of large result sets
- approximate queries / queries from examples