Kostas Patroumpas

Sampling Trajectory Streams with Spatiotemporal Criteria

Monitoring movement of high-dimensional points is essential for environmental databases, geospatial applications, and biodiversity informatics as it reveals crucial information about data evolution, provenance detection, pattern matching etc. Despite recent research interest on processing continuous queries in the context of spatiotemporal data streams, the main focus is on managing the current location of numerous moving objects. In this paper, we turn our attention onto a historical perspective of movement and examine trajectories generated by streaming positional updates. The key challenge is how to maintain a concise, yet quite reliable summary of each objects movement, avoiding any superfluous details and saving in processing complexity and communication cost. We propose two single-pass approximation techniques based on sampling that take advantage of the spatial locality and temporal timeliness inherent in trajectory streams. As a means of reducing substantially the scale of the datasets, we utilize heuristic prediction to distinguish which locations to preserve in the compressed trajectories. A comprehensive experimental study verifies the stability and robustness of the proposed techniques and demonstrates that intelligent compression schemes are able to act as effective load shedding operators achieving remarkable results

Window specification over data streams

Several query languages have been proposed for managing data streams in modern monitoring applications. Continuous queries expressed in these languages usually employ windowing constructs in order to extract finite portions of the potentially unbounded stream. Explicitly or not, window specifications rely on ordering. Usually, timestamps are attached to all tuples flowing into the system as a means to provide ordered access to data items. Several window types have been implemented in stream prototype systems, but a precise definition of their semantics is still lacking. In this paper, we describe a formal framework for expressing windows in continuous queries over data streams. After classifying windows according to their basic characteristics, we give algebraic expressions for the most significant window types commonly appearing in applications. As an essential step towards a stream algebra, we then propose formal definitions for the windowed analogs of typical relational operators, such as join, union or aggregation, and we identify several properties useful to query optimization.

Maintaining consistent results of continuous queries under diverse window specifications

Continuous queries applied over nonterminating data streams usually specify windows in order to obtain an evolving-yet restricted-set of tuples and thus provide timely and incremental results. Although sliding windows get frequently employed in many user requests, additional types like partitioned or landmark windows are also available in stream processing engines. In this paper, we set out to study the existence of monotonic-related semantics for a rich set of windowing constructs in order to facilitate a more efficient maintenance of their changing contents. After laying out a formal foundation for expressing windowed queries, we investigate update patterns observed in most common window variants as well as their impact on adaptations of typical operators (like windowed join, union or aggregation), thus offering more insight towards design and implementation of stream processing mechanisms. Furthermore, we identify syntactic equivalences in algebraic expressions involving windows, to the potential benefit of query optimizations. Finally, this framework is validated for several windowed operations against streaming datasets with simulations at diverse arrival rates and window specifications, providing concrete evidence of its significance.

Online event recognition from moving vessel trajectories

We present a system for online monitoring of maritime activity over streaming positions from numerous vessels sailing at sea. The system employs an online tracking module for detecting important changes in the evolving trajectory of each vessel across time, and thus can incrementally retain concise, yet reliable summaries of its recent movement. In addition, thanks to its complex event recognition module, this system can also offer instant notification to marine authorities regarding emergency situations, such as suspicious moves in protected zones, or package picking at open sea. Not only did our extensive tests validate the performance, efficiency, and robustness of the system against scalable volumes of real-world and synthetically enlarged datasets, but its deployment against online feeds from vessels has also confirmed its capabilities for effective, real-time maritime surveillance.

Towards GeoSpatial semantic data management: strengths, weaknesses, and challenges ahead

An immense wealth of data is already accessible through the Semantic Web and an increasing part of it also has geospatial context or relevance. Although existing technology is mature enough to integrate a variety of information from heterogeneous sources into interlinked features, it still falls behind when it comes to representation and reasoning on spatial characteristics. It is only lately that several RDF stores have begun to accommodate geospatial entities and to enable some kind of processing on them. To address interoperability, the OGC has recently adopted the GeoSPARQL standard, which defines a vocabulary for representing geometric types in RDF and an extension to the SPARQL language for formulating queries. In this paper, we provide a comprehensive review of the current state-of-the-art in geospatially-enabled semantic data management. Apart from an insightful analysis of the available architectures in industry and academia, we conduct an evaluation study on prominent RDF stores with geospatial support. We also compare their performance and attested capabilities to renowned DBMSs widely used in geospatial applications. We introduce a methodology suitable to assess RDF stores for robustness against large geospatial datasets, and also for expressiveness on a variety of queries involving both spatial and thematic criteria. As our findings demonstrate, the potential for query optimization, advanced indexing schemes, and spatio-semantic extensions is significant. Towards this goal, we point out several challenging issues for joint research by the GIS and Semantic Web communities.

Event processing and real-time monitoring over streaming traffic data

Tracking mobility of humans, animals or merchandise has recently given rise to a wide variety of location-based services and monitoring applications. In this paper, we particularly focus on real-time traffic surveillance over densely congested road networks in large metropolitan areas. In such a setting, streaming positional updates are being frequently relayed into a central server from numerous moving vehicles (buses, taxis, passenger cars etc.). Our analysis concerns two important aspects. First, we outline characteristics of a robust processing engine that is capable to efficiently manage such massive, transient, and perhaps noisy geospatial data. Our objective is to provide online aggregates and reliable estimates regarding the current traffic situation at multiple levels of resolution. At a second step, we design a framework for effective multi-modal dissemination of derived information to the end users, in the form of interactive maps for intuitive visualization as well as instant notifications via message feeds. As a proof of concept, we also report on our ongoing development of EPOPS; in its current version, this functional prototype of the proposed scheme is able to deliver cross-platform geographic, textual, and even multimedia content through web and smartphone interfaces.

Multi-granular Time-Based Sliding Windows over Data Streams

We introduce a multi-level window operator that concurrently spans temporal extents of increasing granularity over a streaming dataset. This windowing construct is inherently sliding with time, essentially providing at each granularity a varying, but always finite portion of the most recent stream items. After a careful algebraic formulation of its semantics, we investigate interesting properties and suggest a suitable data structure that can efficiently maintain tuples qualifying for each granular level. Moreover, we propose techniques for evaluating advanced continuous requests against multiple time horizons, achieving near real-time response at reduced overhead. Finally, this framework is empirically validated against streaming data, offering concrete evidence of its benefits to online stream processing.

Monitoring Orientation of Moving Objects around Focal Points

We consider a setting with numerous location-aware moving objects that communicate with a central server. Assuming a set of focal points of interest, we aim at continuously monitoring object orientations and hence detect situations where many objects get closer to or move away from any such site. Towards this goal, we propose a streaming approach that delegates part of the processing to objects, which relay positional updates upon significant deviations at their course. The central processor maintains the changing distribution of current object headings around each focal point and may issue alerts once it observes many objects moving along a direction (e.g., increased northbound traffic near the stadium). To efficiently answer such navigational queries, we introduce a novel access method that indexes object headings influencing a specific site. Furthermore, we extent this scheme to examine trajectory movements around sites over the recent past. Experimental results verify that this framework is able to cope with scalable numbers of objects at reduced communication cost, while offering instant notification of important trends along diverse directions for multiple focal points.

Towards a Flexible and Scalable Fleet Management Service

GPS positioning devices are becoming a commodity sensor platform with the emergence and popularity of smartphones. This abundance of GPS trajectories has fueled significant research around map-matching and related applications such as traffic assessment and prediction. Unfortunately, this research has only been used in costly and complex fleet management solutions. Our latest research endeavor addresses this issue by presenting cost-effective solutions for adapting state-of-the-art research around map-matching and live traffic assessment in the context of fleet management applications. This paper showcases various research results wrapped in a single extensible fleet management platform.

Online amnesic summarization of streaming locations

Massive data streams of positional updates become increasingly difficult to manage under limited memory resources, especially in terms of providing near real-time response to multiple continuous queries. In this paper, we consider online maintenance for spatiotemporal summaries of streaming positions in an aging-aware fashion, by gradually evicting older observations in favor of greater precision for the most recent portions of movement. Although several amnesic functions have been proposed for approximation of time series, we opt for a simple, yet quite efficient scheme that achieves contiguity along all retained stream pieces. To this end, we adapt an amnesic tree structure that effectively meets the requirements of time-decaying approximation while taking advantage of the succession inherent in positional updates. We further exemplify the significance of this scheme in two important cases: the first one refers to trajectory compression of individual objects; the other offers estimated aggregates of moving object locations across time. Both techniques are validated with comprehensive experiments, confirming their suitability in maintaining online concise synopses for moving objects.