@inproceedings {INPROC-2018-45,
author = {Christian Mayer and Ruben Mayer and Sukanya Bhowmik and Lukas Epple and Kurt Rothermel},
title = {{HYPE: Massive Hypergraph Partitioning with Neighborhood Expansion}},
booktitle = {Proceedings of the 2018 IEEE International Conference on Big Data (BigData '18); Seattle, WA, USA, December 10-13, 2018},
publisher = {IEEE},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {1--10},
type = {Konferenz-Beitrag},
month = {Dezember},
year = {2018},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2018-45/INPROC-2018-45.pdf},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {Many important real-world applications---such as social networks or distributed
data bases---can be modeled as hypergraphs. In such a model, vertices represent
entities---such as users or data records---whereas hyperedges model a group
membership of the vertices---such as the authorship in a specific topic or the
membership of a data record in a specific replicated shard. To optimize such
applications, we need an efficient and effective solution to the NP-hard
balanced k-way hypergraph partitioning problem. However, existing hypergraph
partitioners that scale to very large graphs do not effectively exploit the
hypergraph structure when performing the partitioning decisions. We propose
HYPE, a hypergraph partitionier that exploits the neighborhood relations
between vertices in the hypergraph using an efficient implementation of
neighborhood expansion. HYPE improves partitioning quality by up to 95\% and
reduces runtime by up to 39\% compared to the state of the art.},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2018-45&engl=0}
}
@inproceedings {INPROC-2017-72,
author = {Ruben Mayer and Christian Mayer and Larissa Laich},
title = {{The TensorFlow Partitioning and Scheduling Problem: It’s the Critical Path!}},
booktitle = {Proceedings of DIDL'17: Workshop on Distributed Infrastructures for Deep Learning (DIDL'17)},
publisher = {ACM},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {1--6},
type = {Workshop-Beitrag},
month = {Dezember},
year = {2017},
keywords = {TensorFlow; partitioning; scheduling; critical path},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2017-72/INPROC-2017-72.pdf},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2017-72&engl=0}
}
@inproceedings {INPROC-2017-36,
author = {Christian Mayer and Ruben Mayer and Majd Abdo},
title = {{StreamLearner: Distributed Incremental Machine Learning on Event Streams: Grand Challenge}},
booktitle = {Proceedings of the 11th ACM International Conference on Distributed and Event-based Systems},
publisher = {ACM},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {298--303},
type = {Konferenz-Beitrag},
month = {Juni},
year = {2017},
doi = {10.1145/3093742.3095103},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2017-36/INPROC-2017-36.pdf},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2017-36&engl=0}
}
@inproceedings {INPROC-2016-17,
author = {Ruben Mayer and Christian Mayer and Muhammad Adnan Tariq and Kurt Rothermel},
title = {{GraphCEP - Real-time Data Analytics Using Parallel Complex Event and Graph Processing}},
booktitle = {Proceedings of the 10th ACM International Conference on Distributed Event-Based Systems, DEBS'16},
publisher = {ACM},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {1--8},
type = {Konferenz-Beitrag},
month = {Juni},
year = {2016},
doi = {10.1145/2933267.2933509},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2016-17/INPROC-2016-17.pdf},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {In recent years, the proliferation of highly dynamic graph-structured data
streams fueled the demand for real-time data analytics. For instance, detecting
recent trends in social networks enables new applications in areas such as
disaster detection, business analytics or health-care. Parallel Complex Event
Processing has evolved as the paradigm of choice to analyze data streams in a
timely manner, where the incoming data streams are split and processed
independently by parallel operator instances. However, the degree of
parallelism is limited by the feasibility of splitting the data streams into
independent parts such that correctness of event processing is still ensured.
In this paper, we overcome this limitation for graph-structured data by further
parallelizing individual operator instances using modern graph processing
systems. These systems partition the graph data and execute graph algorithms in
a highly parallel fashion, for instance using cloud resources. To this end, we
propose a novel graph-based Complex Event Processing system GraphCEP and
evaluate its performance in the setting of two case studies from the DEBS Grand
Challenge 2016.},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2016-17&engl=0}
}
@inproceedings {INPROC-2016-13,
author = {Christian Mayer and Muhammad Adnan Tariq and Chen Li and Kurt Rothermel},
title = {{GrapH: Heterogeneity-Aware Graph Computation with Adaptive Partitioning}},
booktitle = {Proceedings of the 2016 IEEE 36th International Conference on Distributed Computing Systems (ICDCS)},
publisher = {IEEE},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {118--128},
type = {Konferenz-Beitrag},
month = {Juni},
year = {2016},
doi = {10.1109/ICDCS.2016.92},
issn = {1063-6927},
keywords = {cloud computing; data analysis; GrapH; GraphX; PowerGraph; Pregel; adaptive edge migration strategy; adaptive partitioning; data access locality; data analytics; diverse vertex traffic; expensive network links; graph vertices; graph-structured data; heterogeneity-aware graph computation; heterogeneous network; specialized graph partitioning algorithms; suboptimal partitioning decisions; vertex-centric graph processing; vertex-cut graph partitioning; Automata; Computational modeling; Data analysis; Distributed databases; Heuristic algorithms; Mirrors; Partitioning algorithms},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2016-13/INPROC-2016-13.pdf,
http://dx.doi.org/10.1109/ICDCS.2016.92},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {Vertex-centric graph processing systems such as Pregel, PowerGraph, or GraphX
recently gained popularity due to their superior performance of data analytics
on graph-structured data. These systems exploit the graph structure to improve
data access locality during computation, making use of specialized graph
partitioning algorithms. Recent partitioning techniques assume a uniform and
constant amount of data exchanged between graph vertices (i.e., uniform vertex
traffic) and homogeneous underlying network costs. However, in real-world
scenarios vertex traffic and network costs are heterogeneous. This leads to
suboptimal partitioning decisions and inefficient graph processing. To this
end, we designed GrapH, the first graph processing system using vertex-cut
graph partitioning that considers both, diverse vertex traffic and
heterogeneous network, to minimize overall communication costs. The main idea
is to avoid frequent communication over expensive network links using an
adaptive edge migration strategy. Our evaluations show an improvement of 60\% in
communication costs compared to state-of-the-art partitioning approaches.},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2016-13&engl=0}
}
@inproceedings {INPROC-2015-42,
author = {Thomas Bach and Muhammad Adnan Tariq and Christian Mayer and Kurt Rothermel},
title = {{Utilizing the Hive Mind - How to Manage Knowledge in Fully Distributed Environments}},
booktitle = {OTM 2015 Conferences},
address = {Rhodos},
publisher = {Springer Verlag},
institution = {Universit{\"a}t Stuttgart, Fakult{\"a}t Informatik, Elektrotechnik und Informationstechnik, Germany},
pages = {1--18},
type = {Konferenz-Beitrag},
month = {Oktober},
year = {2015},
keywords = {Knowledge retrieval; Distributed knowledge; Confidence-based indexing; Indexing; Query routing; Knowledge},
language = {Englisch},
cr-category = {C.2.4 Distributed Systems},
ee = {ftp://ftp.informatik.uni-stuttgart.de/pub/library/ncstrl.ustuttgart_fi/INPROC-2015-42/INPROC-2015-42.pdf},
contact = {thomas.bach@ipvs.uni-stuttgart.de},
department = {Universit{\"a}t Stuttgart, Institut f{\"u}r Parallele und Verteilte Systeme, Verteilte Systeme},
abstract = {By 2020, the Internet of Things will consist of 26 Billion connected devices.
All these devices will be collecting an innumerable amount of raw observations,
for example, GPS positions or communication patterns. In order to benefit from
this enormous amount of information, machine learning algorithms are used to
derive knowledge from the gathered observations. This benefit can be increased
further, if the devices are enabled to collaborate by sharing gathered
knowledge. In a massively distributed environment, this is not an easy task, as
the knowledge on each device can be very heterogeneous and based on a different
amount of observations in diverse contexts. In this paper, we propose two
strategies to route a query for specific knowledge to a device that can answer
it with high confidence. To that end, we developed a confidence metric that
takes the number and variance of the observations of a device into account. Our
routing strategies are based on local routing tables that can either be learned
from previous queries over time or actively maintained by interchanging
knowledge models. We evaluated both routing strategies on real world and
synthetic data. Our evaluations show that the knowledge retrieved by the
presented approaches is up to 96.7 \% as accurate as the global optimum.},
url = {http://www2.informatik.uni-stuttgart.de/cgi-bin/NCSTRL/NCSTRL_view.pl?id=INPROC-2015-42&engl=0}
}