Me too! Thanks for organizing this -- István Ráth, PhD Managing Director IncQuery Labs Ltd.
On 2016. October 21. at 15:44:05, Ábel Hegedüs (abel.hegedus@xxxxxxxxxxxxxxxx) wrote:
|
This sender failed our fraud detection checks and may not be who they appear to be. Learn about spoofing |
Feedback |
Hi Tomi,
sounds great, I am really looking forward to the
presentation.
Cheers,
-----
Ábel Hegedüs
IncQuery Labs Ltd.
On 2016. 10. 21.
9:49:50, Tamás Szabó <tamas.szabo@xxxxxxxxx> wrote:
Hey guys,
We have a new colleague at itemis, Thomas Beyhl, who has worked on
a framework for incremental view graph maintenance as part of his
PhD work.
We figured that his work potentially has relevance to the team's
work, and I asked him to give a presentation during one of the
developer meetings.
The proposed time and date is 15:00 on the 10th of November.
Below you find a short bio and an abstract of the
presentation:
Thomas Beyhl was a researcher at the Hasso Plattner Institute in
Potsdam, Germany. He handed in his PhD thesis about „A Framework
for Incremental View Graph Maintenance“ some weeks ago and now is
an employee of itemis AG in Berlin. In his research, he employed
generalized discrimination networks for incremental graph pattern
matching, instead of Rete networks. Generalized discrimination
networks overcome limitations of Rete networks and enable to steer
the tradeoff between time and space for incremental graph pattern
matching.
Abstract: A Framework for Incremental View Graph Maintenance
Nowadays, graphs are employed when relationships between entities
are in the scope of graph queries to avoid performance-critical
join operations of relational data models. Graph queries are used
to query and modify graphs. For that purpose, graph queries employ
graph pattern matching that is NP-complete for subgraph
isomorphism. Graph views can be employed that keep ready answers in
terms of precalculated graph pattern matches for often stated and
complex graph queries to increase query performance. However, such
graph views must be kept consistent with the graphs from which they
are derived.
Existing approaches for incremental graph pattern matching employ
Rete networks and are limited to certain graph pattern matching
languages. However, generalized networks such as Gator networks can
perform better in time and space at the same time.
I describe how to use incremental graph pattern matching as
technique for maintaining graph views. I present a) a modeling
language, which enables to describe generalized discrimination
networks independently from employed graph pattern matching
technologies, b) an annotation mechanism to store graph pattern
matches efficiently and effectively, and c) an incremental
maintenance algorithm for these generalized discrimination
networks. The evaluation shows that a) the maintenance algorithm
scales when the number of graph nodes and edges increases and b)
can perform better in time and space in comparison to Rete
networks.
Cheers,
Tomi
--
Tamás Szabó
Software Engineer
Tel.: +49 711 342 191 0
Fax.: +49 711 342 191 29
Mobil: +49 171 565 416 9
Web: www.itemis.de
Mail: tamas.szabo@xxxxxxxxx
Skype: szabta89
LinkedIn: de.linkedin.com/pub/tamas-szabo/51/610/a12/
itemis AG
Niederlassung Süd
Industriestrasse 6.
70565 Stuttgart
Rechtlicher Hinweis | Legal note:
Amtsgericht Dortmund, HRB 20621
Vorstand | Board: Jens Wagener (Vors.), Wolfgang Neuhaus
Aufsichtsrat | Supervisory board: Dr. Burkhard Igel (Vors.),
Michael Neuhaus, Jennifer Fiorentino
_______________________________________________
viatra-dev mailing list
viatra-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or
unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/viatra-dev
_______________________________________________
viatra-dev mailing list
viatra-dev@xxxxxxxxxxx
To change your delivery options, retrieve your password, or unsubscribe from this list, visit
https://dev.eclipse.org/mailman/listinfo/viatra-dev
|