2012 IEEE Conference on High Performance Extreme Computing 2012
DOI: 10.1109/hpec.2012.6408680
|View full text |Cite
|
Sign up to set email alerts
|

STINGER: High performance data structure for streaming graphs

Abstract: Abstract-The current research focus on "big data" problems highlights the scale and complexity of analytics required and the high rate at which data may be changing. In this paper, we present our high performance, scalable and portable software, Spatio-Temporal Interaction Networks and Graphs Extensible Representation (STINGER), that includes a graph data structure that enables these applications. Key attributes of STINGER are fast insertions, deletions, and updates on semantic graphs with skewed degree distri… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

2
80
0
2

Year Published

2013
2013
2021
2021

Publication Types

Select...
5
3
1

Relationship

2
7

Authors

Journals

citations
Cited by 176 publications
(84 citation statements)
references
References 5 publications
2
80
0
2
Order By: Relevance
“…Adjacencies are stored in semi-dense lists of blocks of edges. The data structure's basic operations are thread-safe, which enables the insertion and deletion of edges and vertices at rates of millions of updates per second on a modern multicore sharedmemory x86 platform [9]. Combining this with a convenient parallel filtering and traversal allows algorithms to easily process and understand the structural changes in the graph over time to ` identify anomalous relationships, actors, and groups.…”
Section: Stingermentioning
confidence: 99%
“…Adjacencies are stored in semi-dense lists of blocks of edges. The data structure's basic operations are thread-safe, which enables the insertion and deletion of edges and vertices at rates of millions of updates per second on a modern multicore sharedmemory x86 platform [9]. Combining this with a convenient parallel filtering and traversal allows algorithms to easily process and understand the structural changes in the graph over time to ` identify anomalous relationships, actors, and groups.…”
Section: Stingermentioning
confidence: 99%
“…where our work focus on distributed cluster environments. It provides a shared memory data structure [13] for large scale dynamic graph processing. A series of dynamic graph algorithms have been developed using this data structure [18], [19].…”
Section: Related Workmentioning
confidence: 99%
“…Existing research has so far focused on static graph processing, While some solutions have explored computation over a sequence of updates condensed into a set of snapshots of static graphs [11], [12], [5], We argue that naively adopting a system designed for static graph processing for real-time incremental computation over dynamic graphs is inefficient. Even though research has been performed on large scale dynamic graph processing [13], [11], developing incremental algorithms for large-scale graphs can be a daunting programming task.…”
Section: Introductionmentioning
confidence: 99%
“…Our distributed architecture and the implementation are for a well established, generic algorithm for clustering which can be applied to any domain, including the Twitter domain as we have demonstrated in this paper. [33] [34] present work on processing streaming graphs in parallel as opposed to the clustering of streaming feature vectors that our work focuses on. Further, [35] presents a simplified clustering algorithm which does not deal with synchronization among the parallel clustering workers.…”
Section: Distributed Clustering Algorithmsmentioning
confidence: 99%