Search

Within the dynamic world of Big Data, traditional systems typically operate in a passive mode, processing and responding to user queries by returning the requested data. However, this methodology falls short of meeting the evolving demands of users who not only wish to analyze data but also to receive proactive updates on topics of interest. To bridge this gap, Big Active Data (BAD) frameworks have been proposed to support extensive data subscriptions and analytics for millions of subscribers. As data volumes and the number of interested users continue to increase, the imperative to optimize BAD systems for enhanced scalability, performance, and efficiency becomes paramount. To this end, this paper introduces three main optimizations, namely: strategic aggregation, intelligent modifications to the query plan, and early result filtering, all aimed at reinforcing a BAD platform's capability to actively manage and efficiently process soaring rates of incoming data and distribute notifications to larger numbers of subscribers.

Hyperparameter optimization is an essential component in many data science pipelines and typically entails exhaustive time and resource-consuming computations in order to explore the combinatorial search space. Similar to this problem, other key operations in data science pipelines exhibit the exact same properties. Important examples are: neural architecture search, where the goal is to identify the best design choices for a neural network, and query cardinality estimation, where given different predicate values for a SQL query the goal is to estimate the size of the output. In this paper, we abstract away those essential components of data science pipelines and we model them as instances of tensor completion, where each variable of the search space corresponds to one mode of the tensor, and the goal is to identify all missing entries of the tensor, corresponding to all combinations of variable values, starting from a very small sample of observed entries. In order to do so, we first conduct a thorough experimental evaluation of existing state-of-the-art tensor completion techniques and introduce domain-inspired adaptations (such as smoothness across the discretized variable space) and an ensemble technique which is able to achieve state-of-the-art performance. We extensively evaluate existing and proposed methods in a number of datasets generated corresponding to (a) hyperparameter optimization for non-neural network models, (b) neural architecture search, and (c) variants of query cardinality estimation, demonstrating the effectiveness of tensor completion as a tool for automating data science pipelines. Furthermore, we release our generated datasets and code in order to provide benchmarks for future work on this topic.

Honeybees are vital for pollination and food production. Among many factors, extreme temperature (e.g., due to climate change) is particularly dangerous for bee health. Anticipating such extremities would allow beekeepers to take early preventive action. Thus, given sensor (temperature) time series data from beehives, how can we find patterns and do forecasting? Forecasting is crucial as it helps spot unexpected behavior and thus issue warnings to the beekeepers. In that case, what are the right models for forecasting? ARIMA, RNNs, or something else? We propose the EBV (Electronic Bee-Veterinarian) method, which has the following desirable properties: (i) principled: it is based on a) diffusion equations from physics and b) control theory for feedback-loop controllers; (ii) effective: it works well on multiple, real-world time sequences, (iii) explainable: it needs only a handful of parameters (e.g., bee strength) that beekeepers can easily understand and trust, and (iv) scalable: it performs linearly in time. We applied our method to multiple real-world time sequences, and found that it yields accurate forecasting (up to 49% improvement in RMSE compared to baselines), and segmentation. Specifically, discontinuities detected by EBV mostly coincide with domain expert's opinions, showcasing our approach's potential and practical feasibility. Moreover, EBV is scalable and fast, taking about 20 minutes on a stock laptop for reconstructing two months of sensor data., Comment: 9 pages, 7 figure, Accepted at 2024 SIAM International Conference on Data Mining (SDM'24)

The prevalence of location tracking systems has resulted in large volumes of spatiotemporal data generated every day. Addressing reachability queries on such datasets is important for a wide range of applications (surveillance, public health, social networks, etc.) A spatiotemporal reachability query identifies whether a physical item (or information etc.) could have been transferred from the source object $O_S$ to the target object $O_T$ during a time interval $I$ (either directly, or through a chain of intermediate transfers). In previous research on spatiotemporal reachability queries, the number of such transfers is not limited, and the weight of a piece of transferred information remains the same. This paper introduces novel reachability queries, which assume a scenario of information decay. Such queries arise when the value of information that travels through the chain of intermediate objects decreases with each transfer. To address such queries efficiently over large spatiotemporal datasets, we introduce the RICCdecay algorithm. Further, the decay scenario leads to an important extension: if there are many different sources of information, the aggregate value of information an object can obtain varies. As a result, we introduce a top-k reachability problem, identifying the k objects with the highest accumulated information. We also present the RICCtopK algorithm that can efficiently compute top-k reachability with transfer decay queries. An experimental evaluation shows the efficiency of the proposed algorithms over previous approaches., Comment: 10 pages

In many Big Data applications today, information needs to be actively shared between systems managed by different organizations. To enable sharing Big Data at scale, developers would have to create dedicated server programs and glue together multiple Big Data systems for scalability. Developing and managing such glued data sharing services requires a significant amount of work from developers. In our prior work, we developed a Big Active Data (BAD) system for enabling Big Data subscriptions and analytics with millions of subscribers. Based on that, we introduce a new mechanism for enabling the sharing of Big Data at scale declaratively so that developers can easily create and provide data sharing services using declarative statements and can benefit from an underlying scalable infrastructure. We show our implementation on top of the BAD system, explain the data sharing data flow among multiple systems, and present a prototype system with experimental results., Comment: 10 pages, 34 figures, to appear on IEEE Big Data - Workshop on Scalable Cloud Data Management

Query Optimization remains an open problem for Big Data Management Systems. Traditional optimizers are cost-based and use statistical estimates of intermediate result cardinalities to assign costs and pick the best plan. However, such estimates tend to become less accurate because of filtering conditions caused either from undetected correlations between multiple predicates local to a single dataset, predicates with query parameters, or predicates involving user-defined functions (UDFs). Consequently, traditional query optimizers tend to ignore or miscalculate those settings, thus leading to suboptimal execution plans. Given the volume of today's data, a suboptimal plan can quickly become very inefficient. In this work, we revisit the old idea of runtime dynamic optimization and adapt it to a shared-nothing distributed database system, AsterixDB. The optimization runs in stages (re-optimization points), starting by first executing all predicates local to a single dataset. The intermediate result created from each stage is used to re-optimize the remaining query. This re-optimization approach avoids inaccurate intermediate result cardinality estimations, thus leading to much better execution plans. While it introduces the overhead for materializing these intermediate results, our experiments show that this overhead is relatively small and it is an acceptable price to pay given the optimization benefits. In fact, our experimental evaluation shows that runtime dynamic optimization leads to much better execution plans as compared to the current default AsterixDB plans as well as to plans produced by static cost-based optimization (i.e. based on the initial dataset statistics) and other state-of-the-art approaches.

Today, data is being actively generated by a variety of devices, services, and applications. Such data is important not only for the information that it contains, but also for its relationships to other data and to interested users. Most existing Big Data systems focus on passively answering queries from users, rather than actively collecting data, processing it, and serving it to users. To satisfy both passive and active requests at scale, users need either to heavily customize an existing passive Big Data system or to glue multiple systems together. Either choice would require significant effort from users and incur additional overhead. In this paper, we present the BAD (Big Active Data) system, which is designed to preserve the merits of passive Big Data systems and introduce new features for actively serving Big Data to users at scale. We show the design and implementation of the BAD system, demonstrate how BAD facilitates providing both passive and active data services, investigate the BAD system's performance at scale, and illustrate the complexities that would result from instead providing BAD-like services with a "glued" system., Comment: 36 pages, 47 figures, submitted to TOCS

Virtually all of today's Big Data systems are passive in nature, responding to queries posted by their users. Instead, we are working to shift Big Data platforms from passive to active. In our view, a Big Active Data (BAD) system should continuously and reliably capture Big Data while enabling timely and automatic delivery of relevant information to a large pool of interested users, as well as supporting retrospective analyses of historical information. While various scalable streaming query engines have been created, their active behavior is limited to a (relatively) small window of the incoming data. To this end we have created a BAD platform that combines ideas and capabilities from both Big Data and Active Data (e.g., Publish/Subscribe, Streaming Engines). It supports complex subscriptions that consider not only newly arrived items but also their relationships to past, stored data. Further, it can provide actionable notifications by enriching the subscription results with other useful data. Our platform extends an existing open-source Big Data Management System, Apache AsterixDB, with an active toolkit. The toolkit contains features to rapidly ingest semistructured data, share execution pipelines among users, manage scaled user data subscriptions, and actively monitor the state of the data to produce individualized information for each user. This paper describes the features and design of our current BAD data platform and demonstrates its ability to scale without sacrificing query capabilities or result individualization., Comment: 30 pages. Accepted by VLDBJ

The wide use of XML for document management and data exchange has created the need to query large repositories of XML data. To efficiently query such large data collections and take advantage of parallelism, we have implemented Apache VXQuery, an open-source scalable XQuery processor. The system builds upon two other open-source frameworks -- Hyracks, a parallel execution engine, and Algebricks, a language agnostic compiler toolbox. Apache VXQuery extends these two frameworks and provides an implementation of the XQuery specifics (data model, data-model dependent functions and optimizations, and a parser). We describe the architecture of Apache VXQuery, its integration with Hyracks and Algebricks, and the XQuery optimization rules applied to the query plan to improve path expression efficiency and to enable query parallelism. An experimental evaluation using a real 500GB dataset with various selection, aggregation and join XML queries shows that Apache VXQuery performs well both in terms of scale-up and speed-up. Our experiments show that it is about 3x faster than Saxon (an open-source and commercial XQuery processor) on a 4-core, single node implementation, and around 2.5x faster than Apache MRQL (a MapReduce-based parallel query processor) on an eight (4-core) node cluster.

AsterixDB is a new, full-function BDMS (Big Data Management System) with a feature set that distinguishes it from other platforms in today's open source Big Data ecosystem. Its features make it well-suited to applications like web data warehousing, social data storage and analysis, and other use cases related to Big Data. AsterixDB has a flexible NoSQL style data model; a query language that supports a wide range of queries; a scalable runtime; partitioned, LSM-based data storage and indexing (including B+-tree, R-tree, and text indexes); support for external as well as natively stored data; a rich set of built-in types; support for fuzzy, spatial, and temporal types and queries; a built-in notion of data feeds for ingestion of data; and transaction support akin to that of a NoSQL store. Development of AsterixDB began in 2009 and led to a mid-2013 initial open source release. This paper is the first complete description of the resulting open source AsterixDB system. Covered herein are the system's data model, its query language, and its software architecture. Also included are a summary of the current status of the project and a first glimpse into how AsterixDB performs when compared to alternative technologies, including a parallel relational DBMS, a popular NoSQL store, and a popular Hadoop-based SQL data analytics platform, for things that both technologies can do. Also included is a brief description of some initial trials that the system has undergone and the lessons learned (and plans laid) based on those early "customer" engagements.

Aggregation has been an important operation since the early days of relational databases. Today's Big Data applications bring further challenges when processing aggregation queries, demanding adaptive aggregation algorithms that can process large volumes of data relative to a potentially limited memory budget (especially in multiuser settings). Despite its importance, the design and evaluation of aggregation algorithms has not received the same attention that other basic operators, such as joins, have received in the literature. As a result, when considering which aggregation algorithm(s) to implement in a new parallel Big Data processing platform (AsterixDB), we faced a lack of "off the shelf" answers that we could simply read about and then implement based on prior performance studies. In this paper we revisit the engineering of efficient local aggregation algorithms for use in Big Data platforms. We discuss the salient implementation details of several candidate algorithms and present an in-depth experimental performance study to guide future Big Data engine developers. We show that the efficient implementation of the aggregation operator for a Big Data platform is non-trivial and that many factors, including memory usage, spilling strategy, and I/O and CPU cost, should be considered. Further, we introduce precise cost models that can help in choosing an appropriate algorithm based on input parameters including memory budget, grouping key cardinality, and data skew., Comment: 25 Pages

Thousands of documents are made available to the users via the web on a daily basis. One of the most extensively studied problems in the context of such document streams is burst identification. Given a term t, a burst is generally exhibited when an unusually high frequency is observed for t. While spatial and temporal burstiness have been studied individually in the past, our work is the first to simultaneously track and measure spatiotemporal term burstiness. In addition, we use the mined burstiness information toward an efficient document-search engine: given a user's query of terms, our engine returns a ranked list of documents discussing influential events with a strong spatiotemporal impact. We demonstrate the efficiency of our methods with an extensive experimental evaluation on real and synthetic datasets., Comment: VLDB2012

The growing amount of XML encoded data exchanged over the Internet increases the importance of XML based publish-subscribe (pub-sub) and content based routing systems. The input in such systems typically consists of a stream of XML documents and a set of user subscriptions expressed as XML queries. The pub-sub system then filters the published documents and passes them to the subscribers. Pub-sub systems are characterized by very high input ratios, therefore the processing time is critical. In this paper we propose a "pure hardware" based solution, which utilizes XPath query blocks on FPGA to solve the filtering problem. By utilizing the high throughput that an FPGA provides for parallel processing, our approach achieves drastically better throughput than the existing software or mixed (hardware/software) architectures. The XPath queries (subscriptions) are translated to regular expressions which are then mapped to FPGA devices. By introducing stacks within the FPGA we are able to express and process a wide range of path queries very efficiently, on a scalable environment. Moreover, the fact that the parser and the filter processing are performed on the same FPGA chip, eliminates expensive communication costs (that a multi-core system would need) thus enabling very fast and efficient pipelining. Our experimental evaluation reveals more than one order of magnitude improvement compared to traditional pub/sub systems., Comment: CIDR 2009