Marc Plantevit

Mining multidimensional and multilevel sequential patterns

Multidimensional databases have been designed to provide decision makers with the necessary tools to help them understand their data. This framework is different from transactional data as the datasets contain huge volumes of historicized and aggregated data defined over a set of dimensions that can be arranged through multiple levels of granularities. Many tools have been proposed to query the data and navigate through the levels of granularity. However, automatic tools are still missing to mine this type of data in order to discover regular specific patterns. In this article, we present a method for mining sequential patterns from multidimensional databases, at the same time taking advantage of the different dimensions and levels of granularity, which is original compared to existing work. The necessary definitions and algorithms are extended from regular sequential patterns to this particular case. Experiments are reported, showing the significance of this approach.

Mining Dominant Patterns in the Sky

Pattern discovery is at the core of numerous data mining tasks. Although many methods focus on efficiency in pattern mining, they still suffer from the problem of choosing a threshold that influences the final extraction result. The goal of our study is to make the results of pattern mining useful from a user-preference point of view. To this end, we integrate into the pattern discovery process the idea of skyline queries in order to mine skyline patterns in a threshold-free manner. Because the skyline patterns satisfy a formal property of dominations, they not only have a global interest but also have semantics that are easily understood by the user. In this work, we first establish theoretical relationships between pattern condensed representations and skyline pattern mining. We also show that it is possible to compute automatically a subset of measures involved in the user query which allows the patterns to be condensed and thus facilitates the computation of the skyline patterns. This forms the basis for a novel approach to mining skyline patterns. We illustrate the efficiency of our approach over several data sets including a use case from chemo informatics and show that small sets of dominant patterns are produced under various measures.

M 2 SP: mining sequential patterns among several dimensions

Mining sequential patterns aims at discovering correlations between events through time. However, even if many works have dealt with sequential pattern mining, none of them considers frequent sequential patterns involving several dimensions in the general case. In this paper, we propose a novel approach, called M2SP, to mine multidimensional sequential patterns. The main originality of our proposition is that we obtain not only intra-pattern sequences but also inter-pattern sequences. Moreover, we consider generalized multidimensional sequential patterns, called jokerized patterns, in which some of the dimension values may not be instanciated. Experiments on synthetic data are reported and show the scalability of our approach.

Mining Graph Topological Patterns: Finding Covariations among Vertex Descriptors

We propose to mine the graph topology of a large attributed graph by finding regularities among vertex descriptors. Such descriptors are of two types: 1) the vertex attributes that convey the information of the vertices themselves and 2) some topological properties used to describe the connectivity of the vertices. These descriptors are mostly of numerical or ordinal types and their similarity can be captured by quantifying their covariation. Mining topological patterns relies on frequent pattern mining and graph topology analysis to reveal the links that exist between the relation encoded by the graph and the vertex attributes. We propose three interestingness measures of topological patterns that differ by the pairs of vertices considered while evaluating up and down co-variations between vertex descriptors. An efficient algorithm that combines search and pruning strategies to look for the most relevant topological patterns is presented. Besides a classical empirical study, we report case studies on four real-life networks showing that our approach provides valuable knowledge.

HYPE: mining hierarchical sequential patterns

Mining data warehouses is still an open problem as few approaches really take the specificities of this framework into account (e.g. multidimensionality, hierarchies, historized data). Multidimensional sequential patterns have been studied but they do not provide any way to handle hierarchies. In this paper, we propose an original sequential pattern extraction method that takes the hierarchies into account. This method extracts more accurate knowledge and extends our preceding M2SP approach. We define the concepts related to our problems as well as the associated algorithms. The results of our experiments confirm the relevance of our proposal.

Cohesive Co-evolution Patterns in Dynamic Attributed Graphs

We focus on the discovery of interesting patterns in dynamic attributed graphs. To this end, we define the novel problem of mining cohesive co-evolution patterns. Briefly speaking, cohesive co-evolution patterns are tri-sets of vertices, timestamps, and signed attributes that describe the local co-evolutions 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 co-evolution patterns in a dynamic graph. Some experiments performed on both synthetic and real-world datasets demonstrate that our algorithm enables to discover relevant patterns in a feasible time.

Combining Sequence and Itemset Mining to Discover Named Entities in Biomedical Texts: A New Type of Pattern

Biomedical named entity recognition (NER) is a challenging problem. In this paper, we show that mining techniques, such as sequential pattern mining and sequential rule mining, can be useful to tackle this problem but present some limitations. We demonstrate and analyse these limitations and introduce a new kind of pattern called LSR pattern that offers an excellent trade-off between the high precision of sequential rules and the high recall of sequential patterns. We formalise the LSR pattern mining problem first. Then we show how LSR patterns enable us to successfully tackle biomedical NER problems. We report experiments carried out on real datasets that underline the relevance of our proposition.

Condensed Representation of Sequential Patterns According to Frequency-Based Measures

Condensed representations of patterns are at the core of many data mining works and there are a lot of contributions handling data described by items. In this paper, we tackle sequential data and we define an exact condensed representation for sequential patterns according to the frequency-based measures. These measures are often used, typically in order to evaluate classification rules. Furthermore, we show how to infer the best patterns according to these measures, i.e., the patterns which maximize them. These patterns are immediately obtained from the condensed representation so that this approach is easily usable in practice. Experiments conducted on various datasets demonstrate the feasibility and the interest of our approach.

Mining Multidimensional Sequential Patterns over Data Streams

Sequential pattern mining is an active field in the domain of knowledge discovery and has been widely studied for over a decade by data mining researchers. More and more, with the constant progress in hardware and software technologies, real-world applications like network monitoring systems or sensor grids generate huge amount of streaming data. This new data model, seen as a potentially infinite and unbounded flow, calls for new real-time sequence mining algorithms that can handle large volume of information with minimal scans. However, current sequence mining approaches fail to take into account the inherent multidimensionality of the streams and all algorithms merely mine correlations between events among only one dimension. Therefore, in this paper, we propose to take multidimensional framework into account in order to detect high-level changes like trends. We show that multidimensional sequential pattern mining over data streams can help detecting interesting high-level variations. We demonstrate with empirical results that our approach is able to extract multidimensional sequential patterns with an approximate support guarantee over data streams.

Trend Mining in Dynamic Attributed Graphs

Many applications see huge demands of discovering important patterns in dynamic attributed graph. In this paper, we introduce the problem of discovering trend sub-graphs 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 real-world dynamic attributed graphs.