Knowledge Graphs: a comprehensive introduction

Hi all,

The following paper is a comprehensive introduction to Knowledge Graphs, 
aimed at newcomers and non-newcomers alike:

 https://arxiv.org/abs/2003.02320

The goal of the paper it to try to gather together a variety of 
perspectives on the issue of Knowledge Graphs. We hope you (or perhaps 
your students) find it interesting/useful!

It is one of the results of a Dagstuhl seminar [1], which was discussed 
a couple of months ago on the list.

The paper was a collaboration effort spanning around a year, with 18 
authors involved, including myself, Eva Blomqvist, Michael Cochez, 
Claudia d'Amato, Gerard de Melo, Claudio Gutierrez, Sabrina Kirrane, 
José Emilio Labra Gayo, Roberto Navigli, Sebastian Neumaier, 
Axel-Cyrille Ngonga Ngomo, Axel Polleres, Sabbir Rashid, Anisa Rula, 
Lukas Schmelzeisen, Juan Sequeda, Steffen Staab, Antoine Zimmermann (in 
CC). The paper has also benefited from discussion with many other 
people, particularly some of the other attendees of the seminar. Many 
thanks to all those who contributed either directly or indirectly!

----------------------------------------------------------------------

<my-2-pesos>
Maybe just to end with a personal "take-away" message for me from the 
seminar and this work ...

I think a key aspect of Knowledge Graphs is that they are being explored 
in parallel by different communities that have not had as much 
cross-over in the past as they perhaps should have. In particular, 
Knowledge Graphs could become a confluence of all sorts of techniques, 
blending queries, ontologies, information extraction, graph algorithms, 
machine learning, etc., all working within a common framework of 
applying a graph abstraction to data. (I think other attendees of the 
seminar came away with the same message [1].)

Academically speaking, I think Knowledge Graphs are interesting in the 
sense that they can become a "confluence" of these areas, raising 
interesting questions about how to combine ontologies and machine 
learning, or how to combine queries with graph algorithms, etc.

Practically speaking, combining these sorts of techniques is becoming 
more and more necessary (maybe even "natural") for solving a wide 
variety of practical problems. To give a small sense of that, I think 
the video "Knowledge Graphs & Deep Learning at YouTube" [2] is a useful 
anecdote, talking about using collaborative filtering, deep learning, 
multimedia, Wikipedia, Wikidata, RDF, SPARQL, entity linking, etc., all 
towards recommending users videos that they might like to see. (One 
thing that's leaves the anecdote incomplete is the omission of 
ontologies, but it's not difficult to see where they might fit in.)
</my-2-pesos>

Best,
Aidan

[1] https://www.dagstuhl.de/en/program/calendar/semhp/?semnr=18371
  - Report by Eva Blomqvist: 
http://blog.liu.se/semanticweb/2018/09/15/dagstuhl-seminar-on-knowledge-graphs/
  - Report by Paul Groth: 
https://thinklinks.wordpress.com/2018/09/18/trip-report-dagstuhl-seminar-on-knowledge-graphs/
  - Report by Juan Sequeda: 
http://www.juansequeda.com/blog/2018/09/18/trip-report-on-knowledge-graph-dagstuhl-seminar/
  - Seminar report: https://drops.dagstuhl.de/opus/volltexte/2019/10328/

[2] https://www.youtube.com/watch?v=D-bTGefJj0A

Received on Tuesday, 10 March 2020 06:02:53 UTC