The flow of data generated from low-cost modern sensing technologies and wireless telecommunication devices enables novel research fields related to the management of this new kind of data and the implementation of appropriate analytics for knowledge extraction. In this work, we investigate how the traditional data cube model is adapted to trajectory warehouses in order to transform raw location data into valuable information. In particular, we focus our research on three issues that are critical to trajectory data warehousing: (a) the trajectory reconstruction procedure that takes place when loading a moving object database with sampled location data originated e.g. from GPS recordings, (b) the ETL procedure that feeds a trajectory data warehouse, and (c) the aggregation of cube measures for OLAP purposes. We provide design solutions for all these issues and we test their applicability and efficiency in real world settings.

Clustering of high dimensional data streams is an important problem in many application domains, a prominent example being network monitoring. Several approaches have been lately proposed for solving independently the different aspects of the problem. There exist methods for clustering over full dimensional streams and methods for finding clusters in subspaces of high dimensional static data. Yet only a few approaches have been proposed so far which tackle both the stream and the high dimensionality aspects of the problem simultaneously. In this work, we propose a new density-based projected clustering algorithm, HDDStream, for high dimensional data streams. Our algorithm summarizes both the data points and the dimensions where these points are grouped together and maintains these summaries online, as new points arrive over time and old points expire due to ageing. Our experimental results illustrate the effectiveness and the efficiency of HDDStream and also demonstrate that it could serve as a trigger for detecting drastic changes in the underlying stream population, like bursts of network attacks.

Trajectory Database (TD) management is a relatively new topic of database research, which has emerged due to the explosion of mobile devices and positioning technologies. Trajectory similarity search forms an important class of queries in TD with applications in trajectory data analysis and spatiotemporal knowledge discovery. In contrast to related works which make use of generic similarity metrics that virtually ignore the temporal dimension, in this paper we introduce a framework consisting of a set of distance operators based on primitive (space and time) as well as derived parameters of trajectories (speed and direction). The novelty of the approach is not only to provide qualitatively different means to query for similar trajectories, but also to support trajectory clustering and classification mining tasks, which definitely imply a way to quantify the distance between two trajectories. For each of the proposed distance operators we devise highly parametric algorithms, the efficiency of which is evaluated through an extensive experimental study using synthetic and real trajectory datasets.

No abstract

Abstract. One of the most important operations involving Data Mining patterns is computing their similarity. In this paper we present a general framework for comparing both simple and complex patterns, i.e., patterns built up from other patterns. Major features of our framework include the notion of structure and measure similarity, the possibility of managing multiple coupling types and aggregation logics, and the recursive definition of similarity for complex patterns.

Decision trees are among the most popular pattern types in data mining due to their intuitive representation. However, little attention has been given on the definition of measures of semantic similarity between decision trees. In this work, we present a general framework for similarity estimation that includes as special cases the estimation of semantic similarity between decision trees, as well as various forms of similarity estimation on classification datasets with respect to different probability distributions defined over the attribute-class space of the datasets. The similarity estimation is based on the partitions induced by the decision trees on the attribute space of the datasets. We use the proposed framework in order to estimate the semantic similarity of decision trees induced from different subsamples of classification datasets; we evaluate its performance with respect to the empirical semantic similarity, which we estimate on the basis of independent hold-out test sets. The availability of similarity measures on decision trees opens a wide range of possibilities for meta-analysis and meta-mining of the data mining results.

Abstract. In this demonstration paper, we present gRecs, a system for group recommendations that follows a collaborative strategy. We enhance recommendations with the notion of support to model the confidence of the recommendations. Moreover, we propose partitioning users into clusters of similar ones. This way, recommendations for users are produced with respect to the preferences of their cluster members without extensively searching for similar users in the whole user base. Finally, we leverage the power of a top-k algorithm for locating the top-k group recommendations.

No abstract