ORIGAMI: A Novel and Effective Approach for Mining Representative Orthogonal Graph Patterns

Chaoji, Vineet; Hasan, Mohammad Al; Salem, Saeed; Besson, J.; Zaki, Mohammed J.

doi:10.1002/sam.10004

Cited by 45 publications

(48 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the next experiment, we show the effectiveness of the sampling approach over traditional graph mining algorithms. We use two large graph datasets: (1) protein-interaction (PI) graphs taken from [7] and (2) cell-graphs [1]. The cell-graphs have class labels based on whether the corresponding graph belongs to a benign (0) or invasive (1) tissue samples; in this dataset there are an equal number of graphs of either class.…”

Section: Sampling Results On Large Graphsmentioning

confidence: 99%

“…For this graph dataset, none of the existing algorithms could finish in 2 full days for 50% support on a dual-core 2.2 GHz machine with 2GB of memory. [7] also reported similar problem with protein interaction network graphs. This motivates the need to find algorithms that can find a small set of interesting and useful patterns instead of attempting to enumerate the entire set of patterns.…”

Section: Introductionmentioning

confidence: 88%

“…There are also algorithms for mining maximal [15] and closed [36] frequent subgraphs. These algorithms are efficient for datasets of small to moderate sized graph (upto few hundred vertices), but they do not scale well for datasets of larger graphs as demonstrated in [7]. The latter defines pattern representativeness and proposes a randomized algorithm called Origami, which mines a representative set of frequent subgraphs instead of the complete set.…”

Section: Related Workmentioning

confidence: 99%

“…Thus it solves the lack of scalability problem. Authors in [7] further argued that the majority of knowledge mining applications for frequent subgraphs do not require the entire frequent pattern-set.…”

Section: Related Workmentioning

confidence: 99%

See 3 more Smart Citations

Output space sampling for graph patterns

2009

Self Cite

View full text Add to dashboard Cite

Recent interest in graph pattern mining has shifted from finding all frequent subgraphs to obtaining a small subset of frequent subgraphs that are representative, discriminative or significant. The main motivation behind that is to cope with the scalability problem that the graph mining algorithms suffer when mining databases of large graphs. Another motivation is to obtain a succinct output set that is informative and useful. In the same spirit, researchers also proposed sampling based algorithms that sample the output space of the frequent patterns to obtain representative subgraphs. In this work, we propose a generic sampling framework that is based on Metropolis-Hastings algorithm to sample the output space of frequent subgraphs. Our experiments on various sampling strategies show the versatility, utility and efficiency of the proposed sampling approach.

show abstract

Section: Sampling Results On Large Graphsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 88%

Section: Related Workmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

See 2 more Smart Citations

Output space sampling for graph patterns

2009

Self Cite

View full text Add to dashboard Cite

show abstract

“…Recently, the randomization idea has also been suggested in the area of pattern mining. Chaoji et al (2008) have introduced a feature construction method that obtains good patterns by sampling under diversity constraints. However, the suggested method requires the user to tune and specify two parameters that control the diversity (orthogonality) and representativeness respectively.…”

Section: Related Workmentioning

confidence: 99%

Effective feature construction by maximum common subgraph sampling

et al. 2010

View full text Add to dashboard Cite

The standard approach to feature construction and predictive learning in molecular datasets is to employ computationally expensive graph mining techniques and to bias the feature search exploration using frequency or correlation measures. These features are then typically employed in predictive models that can be constructed using, for example, SVMs or decision trees. We take a different approach: rather than mining for all optimal local patterns, we extract features from the set of pairwise maximum common subgraphs. The maximum common subgraphs are computed under the block-and-bridge-preserving subgraph isomorphism from the outerplanar examples in polynomial time. We empirically observe a significant increase in predictive performance when using maximum common subgraph features instead of correlated local patterns on 60 benchmark datasets from NCI. Moreover, we show that when we randomly sample the pairs of graphs from which to extract the maximum common subgraphs, we obtain a smaller set of features that still allows the same predictive performance as methods that exhaustively enumerate all possible patterns. The sampling strategy turns out to be a very good compromise between a slight decrease in predictive performance (although still remaining comparable with state-of-the-art methods) and a significant runtime reduction (two orders of magnitude on a popular medium size chemoinformatics dataset). This suggests that maximum common subgraphs are interesting and meaningful features.

show abstract

A possibilistic framework for the detection of terrorism‐related Twitter communities in social media

Moussaoui

Zaghdoud

Akaichi

2018

Concurrency and Computation

View full text Add to dashboard Cite

Summary Since the appearance of social networks, there was a historic increase of data. Unfortunately, terrorists are taking advantage of the easiness of accessing social networks and they have set up profiles to recruit, radicalize, and raise funds. Most of these profiles have pages that exist as well as new recruits to join the terrorist groups, see, and share information. Therefore, there is a potential need for detecting terrorist communities in social networks in order to search for key hints in posts that appear to promote the militants' cause. In order to remedy this problem, we first use a possibilistic‐clustering algorithm that allows more flexibility when assigning a social network profile to clusters (non‐terrorist, terrorist‐sympathizer, terrorist). Then, we introduce a new possibilistic flexible graph mining method to discover similar subgraphs by applying possibilistic similarity rather than using hard structural exact similarity. We experimentally show the efficiency of our possibilistic approach through a detailed process of tweets extract, semantic processing, and classification of the community detection.

show abstract

ORIGAMI: A Novel and Effective Approach for Mining Representative Orthogonal Graph Patterns

Cited by 45 publications

References 17 publications

Output space sampling for graph patterns

Output space sampling for graph patterns

Effective feature construction by maximum common subgraph sampling

A possibilistic framework for the detection of terrorism‐related Twitter communities in social media

Contact Info

Product

Resources

About