In this paper, we address the problem of processing reverse top-k queries in a parallel and distributed setting. Given a database of objects, a set of user preferences, and a query object q, the reverse top-k query returns the subset of user preferences for which the query object belongs to the top-k results. Although recently, the reverse top-k query operator has been studied extensively, its CPU-intensive nature results in prohibitively expensive processing cost, when applied on vast-sized data sets. This limitation motivates us to explore a parallel processing solution, to enable reverse top-k query evaluation over GBs of data in reasonable execution time. To the best of our knowledge, this is the first work that addresses the problem of parallel reverse top-k query processing. We propose a solution to this problem, called DiPaRT, which is based on MapReduce and is provably correct. DiPaRT is empirically evaluated using GB-sized data sets.

In this paper, we propose NoDA, an abstraction layer consisting of spatio-temporal data access operators, which is used to access NoSQL storage engines in a unified way. NoDA alleviates the burden from big data developers of learning the query language of each NoSQL store, and offers a unified view of the underlying NoSQL store. Our approach is inspired by the equivalent paradigm of drivers (such as JDBC) in the relational database world, where the application code is indifferent to the exact underlying database engine. Still, the challenges in the NoSQL world are manifold, because of the lack of standardization in data access. We focus on the specific case of mobility data, and show how spatial and spatio-temporal operators, such as range queries and k-nearest neighbor, are supported in a unified way. Moreover, we present challenges and solutions for supporting spatial and spatio-temporal data in NoSQL stores.

We present a big data framework for the prediction of streaming trajectory data, enriched from other data sources and exploiting mined patterns of trajectories, allowing accurate long-term predictions with low latency. To meet this goal, we follow a multi-step methodology. First, we efficiently compress surveillance data in an online fashion, by constructing trajectory synopses that are spatio-temporally linked with streaming and archival data from a variety of diverse and heterogeneous data sources. The enriched stream of trajectory synopses is stored in a distributed RDF store, supporting data exploration via SPARQL queries. The enriched stream of synopses along with the raw data is consumed by trajectory prediction algorithms that exploit mined patterns from the RDF store, namely medoids of (sub-) trajectory clusters, which prolong the horizon of useful predictions. The framework is extended with offline and online interactive visual analytics tool to facilitate real world analysis in the maritime and the aviation domains.

Abstract We present a system for online composite event recognition over streaming positions of commercial vehicles. Our system employs a data enrichment module, augmenting the mobility data with external information, such as weather data and proximity to points of interest. In addition, the composite event recognition module, based on a highly optimised logic programming implementation of the Event Calculus, consumes the enriched data and identifies activities that are beneficial in fleet management applications. We evaluate our system on large, real-world data from commercial vehicles, and illustrate its efficiency.

Hot spot analysis is the problem of identifying statistically significant spatial clusters from an underlying data set. In this paper, we study the problem of hot spot analysis for massive trajectory data of moving objects, which has many real-life applications in different domains, especially in the analysis of vast repositories of historical traces of spatio-temporal data (cars, vessels, aircrafts). In order to identify hot spots, we propose an approach that relies on the Getis-Ord statistic, which has been used successfully in the past for point data. Since trajectory data is more than just a collection of individual points, we formulate the problem of trajectory hot spot analysis, using the Getis-Ord statistic. We propose a parallel and scalable algorithm for this problem, called THS, which provides an exact solution and can operate on vast-sized data sets. Moreover, we introduce an approximate algorithm (aTHS) that avoids exhaustive computation and trades-off accuracy for efficiency in a controlled manner. In essence, we provide a method that quantifies the maximum induced error in the approximation, in relation with the achieved computational savings. We develop our algorithms in Apache Spark and demonstrate the scalability and efficiency of our approach using a large, historical, real-life trajectory data set of vessels sailing in the Eastern Mediterranean for a period of three years.