Results 1  10
of
21
Learning patterns in the dynamics of biological networks
 In KDD
, 2009
"... Our dynamic graphbased relational mining approach has been developed to learn structural patterns in biological networks as they change over time. The analysis of dynamic networks is important not only to understand life at the systemlevel, but also to discover novel patterns in other structural d ..."
Abstract

Cited by 12 (0 self)
 Add to MetaCart
(Show Context)
Our dynamic graphbased relational mining approach has been developed to learn structural patterns in biological networks as they change over time. The analysis of dynamic networks is important not only to understand life at the systemlevel, but also to discover novel patterns in other structural data. Most current graphbased data mining approaches overlook dynamic features of biological networks, because they are focused on only static graphs. Our approach analyzes a sequence of graphs and discovers rules that capture the changes that occur between pairs of graphs in the sequence. These rules represent the graph rewrite rules that the first graph must go through to be isomorphic to the second graph. Then, our approach feeds the graph rewrite rules into a machine learning system that learns general transformation rules describing the types of changes that occur for a class of dynamic biological networks. The discovered graphrewriting rules show how biological networks change over time, and the transformation rules show the repeated patterns in the structural changes. In this paper, we apply our approach to biological networks to evaluate our approach and to understand how the biosystems change over time. We evaluate our results using coverage and prediction metrics, and compare to biological literature.
LPTA: A Probabilistic Model for Latent Periodic Topic Analysis
"... Abstract—This paper studies the problem of latent periodic topic analysis from timestamped documents. The examples of timestamped documents include news articles, sales records, financial reports, TV programs, and more recently, posts from social media websites such as Flickr, Twitter, and Facebook. ..."
Abstract

Cited by 7 (1 self)
 Add to MetaCart
(Show Context)
Abstract—This paper studies the problem of latent periodic topic analysis from timestamped documents. The examples of timestamped documents include news articles, sales records, financial reports, TV programs, and more recently, posts from social media websites such as Flickr, Twitter, and Facebook. Different from detecting periodic patterns in traditional time series database, we discover the topics of coherent semantics and periodic characteristics where a topic is represented by a distribution of words. We propose a model called LPTA (Latent Periodic Topic Analysis) that exploits the periodicity of the terms as well as term cooccurrences. To show the effectiveness of our model, we collect several representative datasets including Seminar, DBLP and Flickr. The results show that our model can discover the latent periodic topics effectively and leverage the information from both text and time well. Keywordsperiodic topics; topic modeling; I.
Discovering Descriptive Rules in Relational Dynamic Graphs
"... Graph mining methods have become quite popular and a timely challenge is to discover dynamic properties in evolving graphs or networks. We consider the socalled relational dynamic oriented graphs that can be encoded as nary relations with n ≥ 3 and thus represented by Boolean tensors. Two dimensio ..."
Abstract

Cited by 3 (3 self)
 Add to MetaCart
(Show Context)
Graph mining methods have become quite popular and a timely challenge is to discover dynamic properties in evolving graphs or networks. We consider the socalled relational dynamic oriented graphs that can be encoded as nary relations with n ≥ 3 and thus represented by Boolean tensors. Two dimensions are used to encode the graph adjacency matrices and at least one other denotes time. We design the pattern domain of multidimensional association rules, i.e., non trivial extensions of the popular association rules that may involve subsets of any dimensions in their antecedents and their consequents. First, we design new objective interestingness measures for such rules and it leads to different approaches for measuring the rule confidence. Second, we must compute collections of a priori interesting rules. It is considered here as a postprocessing of the closed patterns that can be extracted efficiently from Boolean tensors. We propose optimizations to support both rule extraction scalability and non redundancy. We illustrate the addedvalue of this new data mining task to discover patterns from a reallife relational dynamic graph.
Multidimensional Association Rules in Boolean Tensors ∗
"... Popular data mining methods support knowledge discovery from patterns that hold in binary relations. We study the generalization of association rule mining within arbitrary nary relations and thus Boolean tensorsinsteadofBooleanmatrices. Indeed, manydatasets of interest correspond to relations whos ..."
Abstract

Cited by 3 (2 self)
 Add to MetaCart
(Show Context)
Popular data mining methods support knowledge discovery from patterns that hold in binary relations. We study the generalization of association rule mining within arbitrary nary relations and thus Boolean tensorsinsteadofBooleanmatrices. Indeed, manydatasets of interest correspond to relations whose number of dimensions is greater or equal to 3. However, just a few proposals deal with rule discovery when both the head and the body can involve subsets of any dimensions. A challenging problem is to provide a semantics to such generalized rules by means of objective interestingness measures that have to be carefully designed. Therefore, we discuss the need for different generalizations of the classical confidence measure. We also present the
Mining most frequently changing component in evolving graphs
 WWW
, 2014
"... Many applications see huge demands of finding important changing areas in evolving graphs. In this paper, given a series of snapshots of an evolving graph, we model and develop algorithms to capture the most frequently changing component (MFCC). Motivated by the intuition that the MFCC should capt ..."
Abstract

Cited by 2 (0 self)
 Add to MetaCart
Many applications see huge demands of finding important changing areas in evolving graphs. In this paper, given a series of snapshots of an evolving graph, we model and develop algorithms to capture the most frequently changing component (MFCC). Motivated by the intuition that the MFCC should capture the densest area of changes in an evolving graph, we propose a simple yet effective model. Using only one parameter, users can control tradeoffs between the “density ” of the changes and the size of the detected area. We verify the effectiveness and the efficiency of our approach on real data sets systematically.
Trend Mining in Dynamic Attributed Graphs
"... Abstract. Many applications see huge demands of discovering important patterns in dynamic attributed graph. In this paper, we introduce the problem of discovering trend subgraphs in dynamic attributed graphs. This new kind of pattern relies on the graph structure and the temporal evolution of the a ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract. Many applications see huge demands of discovering important patterns in dynamic attributed graph. In this paper, we introduce the problem of discovering trend subgraphs in dynamic attributed graphs. This new kind of pattern relies on the graph structure and the temporal evolution of the attribute values. Several interestingness measures are introduced to focus on the most relevant patterns with regard to the graph structure, the vertex attributes, and the time. We design an efficient algorithm that benefits from various constraint properties and provide an extensive empirical study from several realworld dynamic attributed graphs. 1
Cohesive Coevolution Patterns in Dynamic Attributed Graphs
"... Abstract. We focus on the discovery of interesting patterns in dynamic attributed graphs. To this end, we define the novel problem of mining cohesive coevolution patterns. Briefly speaking, cohesive coevolution patterns are trisets of vertices, timestamps, and signed attributes that describe the ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract. We focus on the discovery of interesting patterns in dynamic attributed graphs. To this end, we define the novel problem of mining cohesive coevolution patterns. Briefly speaking, cohesive coevolution patterns are trisets of vertices, timestamps, and signed attributes that describe the local coevolutions of similar vertices at several timestamps according to set of signed attributes that express attributes trends. We design the first algorithm to mine the complete set of cohesive coevolution patterns in a dynamic graph. Some experiments performed on both synthetic and realworld datasets demonstrate that our algorithm enables to discover relevant patterns in a feasible time. 1
Probabilistic Paths and Centrality in Time
, 2010
"... Traditionally, graph centrality measures such as betweenness centrality are applied to discrete, static graphs, where binary edges represent the ‘presence’ or ‘absence’ of a relationship. However, when considering the evolution of networks over time, it is more natural to consider interactions at pa ..."
Abstract

Cited by 1 (0 self)
 Add to MetaCart
Traditionally, graph centrality measures such as betweenness centrality are applied to discrete, static graphs, where binary edges represent the ‘presence’ or ‘absence’ of a relationship. However, when considering the evolution of networks over time, it is more natural to consider interactions at particular timesteps as observational evidence of the latent (i.e., hidden) relationships among entities. In this formulation, there is inherent uncertainty about the strength of the underlying relationships and/or whether they are still active at a particular point in time. For example, if we observe an email communication between two people at time t, that indicates they have an active relationship at t, but at time t + k we are less certain the relationship still holds. In this work, we develop a framework to capture this uncertainty, centered around the notion of probabilistic paths. In order to model the effect of relationship uncertainty on network connectivity and its change over time, we formulate a measure of centrality based on most probable paths of communication, rather than shortest paths. In addition to the notion of the relationship strength, we also incorporate uncertainty with regard to the transmission of information using a binomial prior. We show that shortest paths in a unweighted, discrete graph can be formulated using probabilistic paths with a prior and we develop an algorithm to compute the most likely paths in O(VE + V² log V). We demonstrate the effectiveness of our approach by computing probabilistic betweenness centrality over time in the the Enron email dataset.
Discovering InterDimensional Rules in Dynamic Graphs
"... Abstract. Data mining methods that exploit graph/network have become quite popular and a timely challenge is to consider the discovery of dynamic properties in evolving graphs or networks. In this paper, we consider the dynamic oriented graphs that can be encoded as nary relations with n ≥ 3 such t ..."
Abstract

Cited by 1 (1 self)
 Add to MetaCart
(Show Context)
Abstract. Data mining methods that exploit graph/network have become quite popular and a timely challenge is to consider the discovery of dynamic properties in evolving graphs or networks. In this paper, we consider the dynamic oriented graphs that can be encoded as nary relations with n ≥ 3 such that we have a least 3 dimensions: the dimensions of departure (tail) and arrival (head) vertices plus the time dimension. In other terms, it encodes the sequence of adjacency matrices of the graph. In such datasets, we propose a new semantics for interdimensional rules in dynamic graphs. We define rules that may involve subsets of any dimensions in their antecedents and their consequents and we propose the new objective interestingness measure called the exclusive confidence. We introduce a first algorithm for computing such interdimensional rules and we illustrate the addedvalue of exclusive confidence for supporting the discovery of relevant rules from a reallife dynamic graph. 1
Quantifying social network dynamics
 In Proceedings of the 4th Conference on Computational Aspects of Social Networks (CASoN), IEEE Computer Society
, 2012
"... Proceedings of the 4 th ..."
(Show Context)