Results 1  10
of
12
G.: Detecting anomalies in graphs with numeric labels
 In: CIKM 2011
, 2011
"... This paper presentsYagada, an algorithm to search labelled graphs for anomalies using both structural data and numeric attributes. Yagada is explained using several securityrelated examples and validated with experiments on a physical Access Control database. Quantitative analysis shows that in th ..."
Abstract

Cited by 5 (1 self)
 Add to MetaCart
(Show Context)
This paper presentsYagada, an algorithm to search labelled graphs for anomalies using both structural data and numeric attributes. Yagada is explained using several securityrelated examples and validated with experiments on a physical Access Control database. Quantitative analysis shows that in the upper range of anomaly thresholds, Yagada detects twice as many anomalies as the bestperforming numeric discretization algorithm. Qualitative evaluation shows that the detected anomalies are meaningful, representing a combination of structural irregularities and numerical outliers.
WIGM: Discovery of Subgraph Patterns in a Large Weighted Graph
"... Many research areas have begun representing massive data sets as very large graphs. Thus, graph mining has been an active research area in recent years. Most of the graph mining research focuses on mining unweighted graphs. However, weighted graphs are actually more common. The weight on an edge may ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
(Show Context)
Many research areas have begun representing massive data sets as very large graphs. Thus, graph mining has been an active research area in recent years. Most of the graph mining research focuses on mining unweighted graphs. However, weighted graphs are actually more common. The weight on an edge may represent the likelihood or logarithmic transformation of likelihood of the existence of the edge or the strength of an edge, which is common in many biological networks. In this paper, a weighted subgraph pattern model is proposed to capture the importance of a subgraph pattern and our aim is to find these patterns in a large weighted graph. Two related problems are studied in this paper: (1) discovering all patterns with respect to a given minimum weight threshold and (2) finding k patterns with the highest weights. The weighted subgraph patterns do not possess the antimonotonic property and in turn, most of existing subgraph mining methods could not be directly applied. Fortunately, the 1extension property is identified so that a bounded search can be achieved. A novel weighted graph mining algorithm, namely WIGM, is devised based on the 1extension property. Last but not least, real and synthetic data sets are used to show the effectiveness and efficiency of our proposed models and algorithms. 1
Finding Frequent Subgraphs in Longitudinal Social Network Data Using a Weighted Graph Mining Approach
"... Abstract. The mining of social networks entails a high degree of computational complexity. This complexity is exacerbate when considering longitudinal social network data. To address this complexity issue three weighting schemes are proposed in this paper. The fundamental idea is to reduce the compl ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract. The mining of social networks entails a high degree of computational complexity. This complexity is exacerbate when considering longitudinal social network data. To address this complexity issue three weighting schemes are proposed in this paper. The fundamental idea is to reduce the complexity by considering only the most significant nodes and links. The proposed weighting schemes have been incorporated into the weighted variations and extensions of the well established gSpan frequent subgraph mining algorithm. The focus of the work is the cattle movement network found in Great Britain. A complete evaluation of the proposed approaches is presented using this network. In addition, the utility of the discovered patterns is illustrated by constructing a sequential data set to which a sequential mining algorithm can be applied to capturing the changes in “behavior ” represented by a network.
Motif Mining in Weighted Networks
"... Abstract—Unexpectedly frequent subgraphs, known as motifs, can help in characterizing the structure of complex networks. Most of the existing methods for finding motifs are designed for unweighted networks, where only the existence of connection between nodes is considered, and not their strength or ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
(Show Context)
Abstract—Unexpectedly frequent subgraphs, known as motifs, can help in characterizing the structure of complex networks. Most of the existing methods for finding motifs are designed for unweighted networks, where only the existence of connection between nodes is considered, and not their strength or capacity. However, in many real world networks, edges contain more information than just simple node connectivity. In this paper, we propose a new method to incorporate edge weight information in motif mining. We think of a motif as a subgraph that contains unexpected information, and we define a new significance measurement to assess this subgraph exceptionality. The proposed metric embeds the weight distribution in subgraphs and it is based on weight entropy. We use the gtrie data structure to find instances of ksized subgraphs and to calculate its significance score. Following a statistical approach, the random entropy of subgraphs is then calculated, avoiding the time consuming step of random network generation. The discrimination power of the derived motif profile by the proposed method is assessed against the results of the traditional unweighted motifs through a graph classification problem. We use a set of labeled ego networks of coauthorship in the biology and mathematics fields. The new proposed method is shown to be feasible, achieving even slightly better accuracy. Since it does not require the generation of random networks, it is also computationally faster, and because we are able to use the weight information in computing the motif importance, we can avoid converting weighted networks into unweighted ones.
Finding the most descriptive substructures in graphs with discrete and numeric labels
 J INTELL INF SYST
, 2014
"... ..."
(Show Context)
A New Approach To Graph Based Object Classification On Images
, 2015
"... The main idea of graph based image modeling is that the regions of the image, which contain similar properties, are denoted by graph vertices, and the relations between different regions are denoted by graph edges. The vertex and edge attributes usually describe the characteristics of that region ..."
Abstract
 Add to MetaCart
The main idea of graph based image modeling is that the regions of the image, which contain similar properties, are denoted by graph vertices, and the relations between different regions are denoted by graph edges. The vertex and edge attributes usually describe the characteristics of that region and the relation between regions respectively. A simple approach to keep the structural and topological information of an image is to use digital image representation techniques; for instance, quad trees, etc. By modeling images as graphs, the task of image classification becomes one of classifying graphs. Quad trees have been used for representing images in the form of trees. In this paper we propose an algorithm that discovers the subgraphs present in a graph represented by a stream of labeled nodes and edges.Our algorithm is efficient and is easily tuned by the user to produce interesting patterns from various kinds of graph data.
Survey of finding frequent patterns . . .
, 2011
"... Graphs become increasingly important in modeling complicated structures, such as circuits, images, chemical compounds, protein structures, biological networks, social networks, the web, workflows, and XML documents. Many graph search algorithms have been developed in chemical informatics, computer ..."
Abstract
 Add to MetaCart
Graphs become increasingly important in modeling complicated structures, such as circuits, images, chemical compounds, protein structures, biological networks, social networks, the web, workflows, and XML documents. Many graph search algorithms have been developed in chemical informatics, computer vision, video indexing and text retrieval with the increasing demand on the analysis of large amounts of structured data; graph mining has become an active and important theme in data mining.
– Miyamoto Musashi, in The Book of Five Rings. University of Alberta HEAVYWEIGHT PATTERN MINING IN ATTRIBUTED FLOW GRAPHS
"... ”If you master the principles of swordfencing, when you freely beat one man, you beat any man in the world. The spirit of defeating a man is the same for ten million men. The principle of strategy is ..."
Abstract
 Add to MetaCart
(Show Context)
”If you master the principles of swordfencing, when you freely beat one man, you beat any man in the world. The spirit of defeating a man is the same for ten million men. The principle of strategy is
Classification of MRI Brain Scan Data Using Shape Criteria
"... Two mechanisms for classifying Magnetic Resonance Image (MRI) brain scans according to the nature of the corpus callosum are described. The first mechanism uses a hierarchical decomposition approach whereby each MRI scan is decomposed into a hierarchy of “tiles ” which can then be represented as a t ..."
Abstract
 Add to MetaCart
(Show Context)
Two mechanisms for classifying Magnetic Resonance Image (MRI) brain scans according to the nature of the corpus callosum are described. The first mechanism uses a hierarchical decomposition approach whereby each MRI scan is decomposed into a hierarchy of “tiles ” which can then be represented as a tree structure (one tree per scan). A frequent subgraph data mining mechanism is then applied so that subgraphs that occur frequently across the image set are identified. These frequent subgraphs can be viewed as describing a feature space; as such the input images can be translated, according to this feature space, into a set of feature vectors (one per image) to which standard classification techniques can be applied. The second approach uses a time series mechanism to represent the corpus callosum in each image. Using this representation a prelabelled training set was used to define a Case Base (CB) to which Case Based Reasoning (CBR) techniques can be applied so as to classify new cases. Extremely accurate results were obtained with respect to datasets used for evaluation purposes. 1