Alexandros Labrinidis

Big data and its technical challenges

Exploring the inherent technical challenges in realizing the potential of Big Data.

Balancing energy efficiency and quality of aggregate data in sensor networks

Abstract.In-network aggregation has been proposed as one method for reducing energy consumption in sensor networks. In this paper, we explore two ideas related to further reducing energy consumption in the context of in-network aggregation. The first is by influencing the construction of the routing trees for sensor networks with the goal of reducing the size of transmitted data. To this end, we propose a group-aware network configuration method that “clusters” along the same path sensor nodes that belong to the same group. The second idea involves imposing a hierarchy of output filters on the sensor network with the goal of both reducing the size of transmitted data and minimizing the number of transmitted messages. More specifically, we propose a framework to use temporal coherency tolerances in conjunction with in-network aggregation to save energy at the sensor nodes while maintaining specified quality of data. These tolerances are based on user preferences or can be dictated by the network in cases where the network cannot support the current tolerance level. Our framework, called TiNA, works on top of existing in-network aggregation schemes. We evaluate experimentally our proposed schemes in the context of existing in-network aggregation schemes. We present experimental results measuring energy consumption, response time, and quality of data for Group-By queries. Overall, our schemes provide significant energy savings with respect to communication and a negligible drop in quality of data.

TiNA: a scheme for temporal coherency-aware in-network aggregation

This paper presents TiNA, a scheme for minimizing energy consumption in sensor networks by exploiting end-user tolerance to temporal coherency. TiNA utilizes temporal coherency tolerances to both reduce the amount of information transmitted by individual nodes (communication cost dominates power usage in sensor networks), and to improve quality of data when not all sensor readings can be propagated up the network within a given time constraint. TiNA was evaluated against a traditional in-network aggregation scheme with respect to power savings as well as the quality of data for aggregate queries. Preliminary results show that TiNA can reduce power consumption by up to 50% without any loss in the quality of data.

Exploring the tradeoff between performance and data freshness in database-driven Web servers

Abstract.Personalization, advertising, and the sheer volume of online data generate a staggering amount of dynamic Web content. In addition to Web caching, view materialization has been shown to accelerate the generation of dynamic Web content. View materialization is an attractive solution as it decouples the serving of access requests from the handling of updates. In the context of the Web, selecting which views to materialize must be decided online and needs to consider both performance and data freshness, which we refer to as the online view selection problem. In this paper, we define data freshness metrics, provide an adaptive algorithm for the online view selection problem that is based on user-specified data freshness requirements, and present experimental results. Furthermore, we examine alternative metrics for data freshness and extend our proposed algorithm to handle multiple users and alternative definitions of data freshness.

Algorithms and metrics for processing multiple heterogeneous continuous queries

The emergence of monitoring applications has precipitated the need for Data Stream Management Systems (DSMSs), which constantly monitor incoming data feeds (through registered continuous queries), in order to detect events of interest. In this article, we examine the problem of how to schedule multiple Continuous Queries (CQs) in a DSMS to optimize different Quality of Service (QoS) metrics. We show that, unlike traditional online systems, scheduling policies in DSMSs that optimize for average response time will be different from policies that optimize for average slowdown, which is a more appropriate metric to use in the presence of a heterogeneous workload. Towards this, we propose policies to optimize for the average-case performance for both metrics. Additionally, we propose a hybrid scheduling policy that strikes a fine balance between performance and fairness, by looking at both the average- and worst-case performance, for both metrics. We also show how our policies can be adaptive enough to handle the inherent dynamic nature of monitoring applications. Furthermore, we discuss how our policies can be efficiently implemented and extended to exploit sharing in optimized multi-query plans and multi-stream CQs. Finally, we experimentally show using real data that our policies consistently outperform currently used ones.

Balancing performance and data freshness in web database servers

Personalization, advertising, and the sheer volume of online data generate a staggering amount of dynamic web content. In addition to web caching, View Materialization has been shown to accelerate the generation of dynamic web content. View materialization is an attractive solution as it decouples the serving of access requests from the handling of updates. In the context of the Web, selecting which views to materialize must be decided online and needs to consider both performance and data freshness, which we refer to as the Online View Selection problem. In this paper, we define data freshness metrics, provide an adaptive algorithm for the online view selection problem, and present experimental results.

Global Transcriptional Response to Spermine, a Component of the Intramacrophage Environment, Reveals Regulation of Francisella Gene Expression through Insertion Sequence Elements

Tularemia is caused by the category A biodefense agent Francisella tularensis. This bacterium is associated with diverse environments and a plethora of arthropod and mammalian hosts. How F. tularensis adapts to these different conditions, particularly the eukaryotic intracellular environment in which it replicates, is poorly understood. Here, we demonstrate that the polyamines spermine and spermidine are environmental signals that alter bacterial stimulation of host cells. Genomewide analysis showed that F. tularensis LVS undergoes considerable changes in gene expression in response to spermine. Unexpectedly, analysis of gene expression showed that multiple members of two classes of Francisella insertion sequence (IS) elements, ISFtu1 and ISFtu2, and the genes adjacent to these elements were induced by spermine. Spermine was sufficient to activate transcription of these IS elements and of nearby genes in broth culture and in macrophages. Importantly, the virulent strain of F. tularensis, Schu S4, exhibited similar phenotypes of cytokine induction and gene regulation in response to spermine. Distinctions in gene expression changes between Schu S4 and LVS at one orthologous locus, however, correlated with differences in IS element location. Our results indicate that spermine and spermidine are novel triggers to alert F. tularensis of its eukaryotic host environment. The results reported here also identify an unexpected mechanism of gene regulation controlled by a spermine-responsive promoter contained within IS elements. Different arrangements of these mobile genetic elements among Francisella strains may contribute to virulence by conveying new expression patterns for genes from different strains.

UNIT: User-centric Transaction Management in Web-Database Systems

Web-database systems are nowadays an integral part of everybody’s life, with applications ranging from monitoring/ trading stock portfolios, to personalized blog aggregation and news services, to personalized weather tracking services. For most of these services to be successful (and their users to be kept satisfied), two criteria need to be met: user requests must be answered in a timely fashion and using fresh data. This paper presents a framework to balance both requirements from the users’ perspective. Toward this, we propose a user satisfaction metric to measure the overall effectiveness of the Web-database system. We also provide a set of algorithms to dynamically optimize this metric, through query admission control and update frequency modulation. Finally, we present extensive experimental results which compare our proposed algorithms to the current state of the art and show that we outperform competitors under various workloads (generated based on real traces) and user requirements.

Preemptive rate-based operator scheduling in a data stream management system

Summary form only given. Data stream management systems are being developed to process continuous queries over multiple data streams. These continuous queries are typically used for monitoring purposes where the detection of an event might trigger a sequence of actions or the execution of a set of specified tasks. Such events are identified by tuples produced by a query and hence, it is important to produce the available portions of a query result as early as possible. A core element for improving the interactive performance of a continuous query is the operator scheduler. An operator scheduler is particularly important when the processing requirements and the productivity of different streams are highly skewed. The need for an operator scheduler becomes even more crucial when tuples from different streams arrive asynchronously. To meet these needs, we are proposing a preemptive rate-based scheduling policy that handles the asynchronous nature of tuple arrival and the heterogeneity in the query plan. Experimental results show the significant improvements provided by our proposed policy.

Caching and Materialization for Web Databases

Database systems have been driving dynamic websites since the early 1990s; nowadays, even seemingly static websites employ a database back-end for personalization and advertising purposes. In order to keep up with the high demand fuelled by the rapid growth of the Internet, a number of caching and materialization techniques have been proposed for web databases over the years. The main goal of these techniques is to improve performance, scalability, and manageability of database-driven dynamic websites, in a way that the quality of data is not compromised. Although caching and materialization are well-understood concepts in the traditional database and networking/operating systems literature, the Web and web databases bring forth unique characteristics that warrant new techniques and approaches. In this monograph, we adopt a data management point of view to describe the system architectures of web databases, and analyze the research issues related to caching and materialization in such architectures. We also present the state-of-the-art in caching and materialization for web databases and organize current approaches according to the fundamental questions, namely how to store, how to use, and how to maintain cached/materialized web data. Finally, we associate work in caching and materialization for web databases to similar techniques in other related areas, such as data warehousing, distributed systems, and distributed databases.