Q-error -- the standard metric for quantifying the error of individual cardinality estimates -- has been widely adopted as a surrogate for query plan optimality in recent work on learning-based cardinality estimation. However, the only result connecting Q-error with plan optimality is an upper-bound on the cost of the worst possible query plan computed from a set of cardinality estimates---there is no connection between Q-error and the real plans generated by standard query optimizers. Therefore, in order to identify sub-optimal query plans, we propose a learning-based method having as its main feature a novel measure called L1-error. Similar to Q-error, L1-error requires complete knowledge of true cardinalities and estimates for all the sub-plans of a query plan. Unlike Q-error, which considers the estimates independently, L1-error is defined as a permutation distance between true cardinalities and estimates for all the sub-plans having the same number of joins. Moreover, L1-error takes into account errors relative to the magnitude of their cardinalities and gives larger weight to small multi-way joins. Our experimental results confirm that, when L1-error is integrated into a standard decision tree classifier, it leads to the accurate identification of sub-optimal plans across four different benchmarks. This accuracy can be further improved by combining L1-error with Q-error into a composite feature that can be computed without overhead from the same data.

Sketches are single-pass small-space data summaries that can quickly estimate the cardinality of join queries. However, sketches are not directly applicable to join queries with dynamic filter conditions --- where arbitrary selection predicate(s) are applied --- since a sketch is limited to a fixed selection. While multiple sketches for various selections can be used in combination, they each incur individual storage and maintenance costs. Alternatively, exact sketches can be built during runtime for every selection. To make this process scale, a high-degree of parallelism --- available in hardware accelerators such as GPUs --- is required. Therefore, sketch usage for cardinality estimation in query optimization is limited. Following recent work that applies transformers to cardinality estimation, we design a novel learning-based method to approximate the sketch of any arbitrary selection, enabling sketches for join queries with filter conditions. We train a transformer on each table to estimate the sketch of any subset of the table, i.e., any arbitrary selection. Transformers achieve this by learning the joint distribution amongst table attributes, which is equivalent to a multidimensional sketch. Subsequently, transformers can approximate any sketch, enabling sketches for join cardinality estimation. In turn, estimating joins via approximate sketches allows tables to be modeled individually and thus scales linearly with the number of tables. We evaluate the accuracy and efficacy of approximate sketches on queries with selection predicates consisting of conjunctions of point and range conditions. Approximate sketches achieve similar accuracy to exact sketches with at least one order of magnitude less overhead.

Cost-based query optimization remains a critical task in relational databases even after decades of research and industrial development. Query optimizers rely on a large range of statistical synopses for accurate cardinality estimation. As the complexity of selections and the number of join predicates increase, two problems arise. First, statistics cannot be incrementally composed to effectively estimate the cost of the sub-plans generated in plan enumeration. Second, small errors are propagated exponentially through joins, which can lead to severely sub-optimal plans. In this paper, we introduce COMPASS, a novel query optimization paradigm for in-memory databases based on a single type of statistics---Fast-AGMS sketches. In COMPASS, query optimization and execution are intertwined. Selection predicates and sketch updates are pushed-down and evaluated online during query optimization. This allows Fast-AGMS sketches to be computed only over the relevant tuples---which enhances cardinality estimation accuracy. Plan enumeration is performed over the query join graph by incrementally composing attribute-level sketches---not by building a separate sketch for every sub-plan. We prototype COMPASS in MapD -- an open-source parallel database -- and perform extensive experiments over the complete JOB benchmark. The results prove that COMPASS generates better execution plans -- both in terms of cardinality and runtime -- compared to four other database systems. Overall, COMPASS achieves a speedup ranging from 1.35X to 11.28X in cumulative query execution time over the considered competitors.

Consider a set of black-box models – each of them independently trained on a different dataset – answering the same predictive spatio-temporal query. Being built in isolation, each model traverses its own life-cycle until it is deployed to production, learning data patterns from different datasets and facing independent hyper-parameter tuning. In order to answer the query, the set of black-box predictors has to be ensembled and allocated to the spatio-temporal query region. However, computing an optimal ensemble is a complex task that involves selecting the appropriate models and defining an effective allocation strategy that maps the models to the query region. In this paper we present DJEnsemble, a cost-based strategy for the automatic selection and allocation of a disjoint ensemble of black-box predictors to answer predictive spatio-temporal queries. We conduct a set of extensive experiments that evaluate DJEnsemble and highlight its efficiency, selecting model ensembles that are almost as efficient as the optimal solution. When compared against the traditional ensemble approach, DJEnsemble achieves up to 4X improvement in execution time and almost 9X improvement in prediction accuracy.

Coordination failure reduces match quality among employers and candidates in the job market, resulting in a large number of unfilled positions and/or unstable, short-term employment. Centralized job search engines provide a platform that connects directly employers with job-seekers. However, they require users to disclose a significant amount of personal data, i.e., build a user profile, in order to provide meaningful recommendations. In this paper, we present PrivateJobMatch - a privacy-oriented deferred multi-match recommender system - which generates stable pairings while requiring users to provide only a partial ranking of their preferences. PrivateJobMatch explores a series of adaptations of the game-theoretic Gale-Shapley deferred acceptance algorithm which combine the flexibility of decentralized markets with the intelligence of centralized matching. We identify the shortcomings of the original algorithm when applied to a job market and propose novel solutions that rely on machine learning techniques. Experimental results on real and synthetic data confirm the benefits of the proposed algorithms across several quality measures. Over the past year, we have implemented a PrivateJobMatch prototype and deployed it in an active job market economy. Using the gathered real-user preference data, we find that the match recommendations are superior to a typical decentralized job market---while requiring only a partial ranking of the user preferences.

As applications continue to generate multi-dimensional data at exponentially increasing rates, fast analytics to extract meaningful results is becoming extremely important. The database community has developed array databases that alleviate this problem through a series of techniques. In-situ mechanisms provide direct access to raw data in the original format---without loading and partitioning. Parallel processing scales to the largest datasets. In-memory caching reduces latency when the same data are accessed across a workload of queries. However, we are not aware of any work on distributed caching of multi-dimensional raw arrays. In this paper, we introduce a distributed framework for cost-based caching of multi-dimensional arrays in native format. Given a set of files that contain portions of an array and an online query workload, the framework computes an effective caching plan in two stages. First, the plan identifies the cells to be cached locally from each of the input files by continuously refining an evolving R-tree index. In the second stage, an optimal assignment of cells to nodes that collocates dependent cells in order to minimize the overall data transfer is determined. We design cache eviction and placement heuristic algorithms that consider the historical query workload. A thorough experimental evaluation over two real datasets in three file formats confirms the superiority - by as much as two orders of magnitude - of the proposed framework over existing techniques in terms of cache overhead and workload execution time.

Science applications are producing an ever-increasing volume of multi-dimensional data that are mainly processed with distributed array databases. These raw arrays are ``cooked'' into derived data products using complex pipelines that are time-consuming. As a result, derived data products are released infrequently and become stale soon thereafter. In this paper, we introduce materialized array views as a database construct for scientific data products. We model the ``cooking'' process as incremental view maintenance with batch updates and give a three-stage heuristic that finds effective update plans. Moreover, the heuristic repartitions the array and the view continuously based on a window of past updates as a side-effect of view maintenance without overhead. We design an analytical cost model for integrating materialized array views in queries. A thorough experimental evaluation confirms that the proposed techniques are able to incrementally maintain a real astronomical data product in a production environment.

In-situ processing has been proposed as a novel data exploration solution in many domains generating massive amounts of raw data, e.g., astronomy, since it provides immediate SQL querying over raw files. The performance of in-situ processing across a query workload is, however, limited by the speed of full scan, tokenizing, and parsing of the entire data. Online aggregation (OLA) has been introduced as an efficient method for data exploration that identifies uninteresting patterns faster by continuously estimating the result of a computation during the actual processing---the computation can be stopped as early as the estimate is accurate enough to be deemed uninteresting. However, existing OLA solutions have a high upfront cost of randomly shuffling and/or sampling the data.

Big Model analytics tackles the training of massive models that go beyond the available memory of a single computing device, e.g., CPU or GPU. It generalizes Big Data analytics which is targeted at how to train memory-resident models over out-of-memory training data. In this paper, we propose an in-database solution for Big Model analytics. We identify dot-product as the primary operation for training generalized linear models and introduce the first array-relation dot-product join database operator between a set of sparse arrays and a dense relation. This is a constrained formulation of the extensively studied sparse matrix vector multiplication (SpMV) kernel. The paramount challenge in designing the dot-product join operator is how to optimally schedule access to the dense relation based on the non-contiguous entries in the sparse arrays. We propose a practical solution characterized by two technical contributions---dynamic batch processing and array reordering. We devise three heuristics -- LSH, Radix, and K-center -- for array reordering and analyze them thoroughly. We execute extensive experiments over synthetic and real data that confirm the minimal overhead the operator incurs when sufficient memory is available and the graceful degradation it suffers as memory becomes scarce. Moreover, dot-product join achieves an order of magnitude reduction in execution time over alternative solutions.

Existing data analytics systems have approached predictive model training exclusively from a data-parallel perspective. Data examples are partitioned to multiple workers and training is executed concurrently over different partitions, under various synchronization policies that emphasize speedup or convergence. Since models with millions and even billions of features become increasingly common nowadays, model management becomes an equally important task for effective training. In this paper, we present a general framework for parallelizing stochastic optimization algorithms over massive models that cannot fit in memory. We extend the lock-free HOGWILD!-family of algorithms to disk-resident models by vertically partitioning the model offline and asynchronously updating the resulting partitions online. Unlike HOGWILD!, concurrent requests to the common model are minimized by a preemptive push-based sharing mechanism that reduces the number of disk accesses. Experimental results on real and synthetic datasets show that the proposed framework achieves improved convergence over HOGWILD! and is the only solution scalable to massive models.

Scientific applications are generating an ever-increasing volume of multi-dimensional data that are largely processed inside distributed array databases and frameworks. Similarity join is a fundamental operation across scientific workloads that requires complex processing over an unbounded number of pairs of multi-dimensional points. In this paper, we introduce a novel distributed similarity join operator for multi-dimensional arrays. Unlike immediate extensions to array join and relational similarity join, the proposed operator minimizes the overall data transfer and network congestion while providing load-balancing, without completely repartitioning and replicating the input arrays. We define formally array similarity join and present the design, optimization strategies, and evaluation of the first array similarity join operator.

Traditional databases are not equipped with the adequate functionality to handle the volume and variety of "Big Data". Strict schema definition and data loading are prerequisites even for the most primitive query session. Raw data processing has been proposed as a schema-on-demand alternative that provides instant access to the data. When loading is an option, it is driven exclusively by the current-running query, resulting in sub-optimal performance across a query workload. In this paper, we investigate the problem of workload-driven raw data processing with partial loading. We model loading as fully-replicated binary vertical partitioning. We provide a linear mixed integer programming optimization formulation that we prove to be NP-hard. We design a two-stage heuristic that comes within close range of the optimal solution in a fraction of the time. We extend the optimization formulation and the heuristic to pipelined raw data processing, scenario in which data access and extraction are executed concurrently. We provide three case-studies over real data formats that confirm the accuracy of the model when implemented in a state-of-the-art pipelined operator for raw data processing.

Traditional databases incur a significant data-to-query delay due to the requirement to load data inside the system before querying. Since this is not acceptable in many domains generating massive amounts of raw data (e.g., genomics), databases are entirely discarded. External tables , on the other hand, provide instant SQL querying over raw files. Their performance across a query workload is limited though by the speed of repeated full scans, tokenizing, and parsing of the entire file. In this article, we propose SCANRAW, a novel database meta-operator for in-situ processing over raw files that integrates data loading and external tables seamlessly, while preserving their advantages: optimal performance across a query workload and zero time-to-query. We decompose loading and external table processing into atomic stages in order to identify common functionality. We analyze alternative implementations and discuss possible optimizations for each stage. Our major contribution is a parallel superscalar pipeline implementation that allows SCANRAW to take advantage of the current many- and multicore processors by overlapping the execution of independent stages. Moreover, SCANRAW overlaps query processing with loading by speculatively using the additional I/O bandwidth arising during the conversion process for storing data into the database, such that subsequent queries execute faster. As a result, SCANRAW makes intelligent use of the available system resources—CPU cycles and I/O bandwidth—by switching dynamically between tasks to ensure that optimal performance is achieved. We implement SCANRAW in a state-of-the-art database system and evaluate its performance across a variety of synthetic and real-world datasets. Our results show that SCANRAW with speculative loading achieves the best-possible performance for a query sequence at any point in the processing. Moreover, SCANRAW maximizes resource utilization for the entire workload execution while speculatively loading data and without interfering with normal query processing.

Antijoin cardinality estimation is among a handful of problems that has eluded accurate efficient solutions amenable to implementation in relational query optimizers. Given the widespread use of antijoin and subset-based queries in analytical workloads and the extensive research targeted at join cardinality estimation—a seemingly related problem—the lack of adequate solutions for antijoin cardinality estimation is intriguing. In this article, we introduce a novel sampling-based estimator for antijoin cardinality that (unlike existent estimators) provides sufficient accuracy and efficiency to be implemented in a query optimizer. The proposed estimator incorporates three novel ideas. First, we use prior workload information when learning a mixture superpopulation model of the data offline. Second, we design a Bayesian statistics framework that updates the superpopulation model according to the live queries, thus allowing the estimator to adapt dynamically to the online workload. Third, we develop an efficient algorithm for sampling from a hypergeometric distribution in order to generate Monte Carlo trials , without explicitly instantiating either the population or the sample. When put together, these ideas form the basis of an efficient antijoin cardinality estimator satisfying the strict requirements of a query optimizer, as shown by the extensive experimental results over synthetically-generated as well as massive TPC-H data.

Traditional databases incur a significant data-to-query delay due to the requirement to load data inside the system before querying. Since this is not acceptable in many domains generating massive amounts of raw data, e.g., genomics, databases are entirely discarded. External tables, on the other hand, provide instant SQL querying over raw files. Their performance across a query workload is limited though by the speed of repeated full scans, tokenizing, and parsing of the entire file. In this paper, we propose SCANRAW, a novel database physical operator for in-situ processing over raw files that integrates data loading and external tables seamlessly while preserving their advantages: optimal performance across a query workload and zero time-to-query. Our major contribution is a parallel super-scalar pipeline implementation that allows SCANRAW to take advantage of the current many- and multi-core processors by overlapping the execution of independent stages. Moreover, SCANRAW overlaps query processing with loading by speculatively using the additional I/O bandwidth arising during the conversion process for storing data into the database such that subsequent queries execute faster. As a result, SCANRAW makes optimal use of the available system resources -- CPU cycles and I/O bandwidth -- by switching dynamically between tasks to ensure that optimal performance is achieved. We implement SCANRAW in a state-of-the-art database system and evaluate its performance across a variety of synthetic and real-world datasets. Our results show that SCANRAW with speculative loading achieves optimal performance for a query sequence at any point in the processing. Moreover, SCANRAW maximizes resource utilization for the entire workload execution while speculatively loading data and without interfering with normal query processing.

Scientific data have dual structure. Raw data are preponderantly ordered multi-dimensional arrays or sequences while metadata and derived data are best represented as unordered relations. Scientific data processing requires complex operations over arrays and relations. These operations cannot be expressed using only standard linear and relational algebra operators, respectively. Existing scientific data processing systems are designed for a single data model and handle complex processing at the application level.

Online aggregation provides continuous estimates to the final result of a computation during the actual processing. The user can stop the computation as soon as the estimate is accurate enough, typically early in the execution, or can let the processing terminate and obtain the exact result. In this demonstration, we introduce a general framework for parallel online aggregation in which estimation does not incur overhead on top of the actual processing. We define a generic interface to express any estimation model that abstracts completely the execution details. We design multiple sampling-based estimators suited for parallel online aggregation and implement them inside the framework. Demonstration participants are shown how estimates to general SQL aggregation queries over terabytes of TPC-H data are generated during the entire processing. Due to parallel execution, the estimate converges to the correct result in a matter of seconds even for the most difficult queries. The behavior of the estimators is evaluated under different operating regimes of the distributed cluster used in the demonstration.

We present GLADE, a scalable distributed system for large scale data analytics. GLADE takes analytical functions expressed through the User-Defined Aggregate (UDA) interface and executes them efficiently on the input data. The entire computation is encapsulated in a single class which requires the definition of four methods. The runtime takes the user code and executes it right near the data by taking full advantage of the parallelism available inside a single machine as well as across a cluster of computing nodes.

Sampling is used as a universal method to reduce the running time of computations - the computation is performed on a much smaller sample and then the result is scaled to compensate for the difference in size. Sketches are a popular approximation method for data streams and they proved to be useful for estimating frequency moments and aggregates over joins. A possibility to further improve the time performance of sketches is to compute the sketch over a sample of the stream rather than the entire data stream. In this paper we analyze the behavior of the sketch estimator when computed over a sample of the stream, not the entire data stream, for the size of join and the self-join size problems. Our analysis is developed for a generic sampling process. We instantiate the results of the analysis for all three major types of sampling - Bernoulli sampling which is used for load shedding, sampling with replacement which is used to generate i.i.d. samples from a distribution, and sampling without replacement which is used by online aggregation engines - and compare these particular results with the results of the basic sketch estimator. Our experimental results show that the accuracy of the sketch computed over a small sample of the data is, in general, close to the accuracy of the sketch estimator computed over the entire data even when the sample size is only 10% or less of the dataset size. This is equivalent to a speed-up factor of at least 10 when updating the sketch.

DBO is a database system that utilizes randomized algorithms to give statistically meaningful estimates for the final answer to a multi-table, disk-based query from start to finish during query execution. However, DBO's "time 'til utility" (or "TTU"; that is, the time until DBO can give a useful estimate) can be overly large, particularly in the case that many database tables are joined in a query, or in the case that a join query includes a very selective predicate on one or more of the tables, or when the data are skewed. In this paper, we describe Turbo DBO , which is a prototype database system that can answer multi-table join queries in a scalable fashion, just like DBO. However, Turbo DBO often has a much lower TTU than DBO. The key innovation of Turbo DBO is that it makes use of novel algorithms that look for and remember "partial match" tuples in a randomized fashion. These are tuples that satisfy some of the boolean predicates associated with the query, and can possibly be grown into tuples that actually contribute to the final query result at a later time.

We demonstrate our prototype of the DBO database system. DBO is designed to facilitate scalable analytic processing over large data archives. DBO's analytic processing performance is competitive with other database systems; however, unlike any other existing research or industrial system, DBO maintains a statistically meaningful guess to the final answer to a query from start to finish during query processing. This guess may be quite accurate after only a few seconds or minutes, while answering a query exactly may take hours. This can result in significant savings in both user and computer time, since a user can abort a query as soon as he or she is happy with the guess' accuracy.

Sketching techniques provide approximate answers to aggregate queries both for data-streaming and distributed computation. Small space summaries that have linearity properties are required for both types of applications. The prevalent method for analyzing sketches uses moment analysis and distribution-independent bounds based on moments. This method produces clean, easy to interpret, theoretical bounds that are especially useful for deriving asymptotic results. However, the theoretical bounds obscure fine details of the behavior of various sketches and they are mostly not indicative of which type of sketches should be used in practice. Moreover, no significant empirical comparison between various sketching techniques has been published, which makes the choice even harder. In this article we take a close look at the sketching techniques proposed in the literature from a statistical point of view with the goal of determining properties that indicate the actual behavior and producing tighter confidence bounds. Interestingly, the statistical analysis reveals that two of the techniques, Fast-AGMS and Count-Min, provide results that are in some cases orders of magnitude better than the corresponding theoretical predictions. We conduct an extensive empirical study that compares the different sketching techniques in order to corroborate the statistical analysis with the conclusions we draw from it. The study indicates the expected performance of various sketches, which is crucial if the techniques are to be used by practitioners. The overall conclusion of the study is that Fast-AGMS sketches are, for the full spectrum of problems, either the best, or close to the best, sketching technique. We apply the insights obtained from the statistical study and the experimental results to design effective algorithms for sketching interval data. We show how the two basic methods for sketching interval data, DMAP and fast range-summation, can be improved significantly with respect to the update time without a significant loss in accuracy. The gain in update time can be as large as two orders of magnitude, thus making the improved methods practical. The empirical study suggests that DMAP is preferable when update time is the critical requirement and fast range-summation is desirable for better accuracy.

Sketching techniques can provide approximate answers to aggregate queries either for data-streaming or distributed computation. Small space summaries that have linearity properties are required for both types of applications. The prevalent method for analyzing sketches uses moment analysis and distribution independent bounds based on moments. This method produces clean, easy to interpret, theoretical bounds that are especially useful for deriving asymptotic results. However, the theoretical bounds obscure fine details of the behavior of various sketches and they are mostly not indicative of which type of sketches should be used in practice. Moreover, no significant empirical comparison between various sketching techniques has been published, which makes the choice even harder. In this paper, we take a close look at the sketching techniques proposed in the literature from a statistical point of view with the goal of determining properties that indicate the actual behavior and producing tighter confidence bounds. Interestingly, the statistical analysis reveals that two of the techniques, Fast-AGMS and Count-Min, provide results that are in some cases orders of magnitude better than the corresponding theoretical predictions. We conduct an extensive empirical study that compares the different sketching techniques in order to corroborate the statistical analysis with the conclusions we draw from it. The study indicates the expected performance of various sketches, which is crucial if the techniques are to be used by practitioners. The overall conclusion of the study is that Fast-AGMS sketches are, for the full spectrum of problems, either the best, or close to the best, sketching technique. This makes Fast-AGMS sketches the preferred choice irrespective of the situation.

The exact computation of aggregate queries, like the size of join of two relations, usually requires large amounts of memory (constrained in data-streaming) or communication (constrained in distributed computation) and large processing times. In this situation, approximation techniques with provable guarantees, like sketches, are one possible solution. The performance of sketches depends crucially on the ability to generate particular pseudo-random numbers. In this article we investigate both theoretically and empirically the problem of generating k -wise independent pseudo-random numbers and, in particular, that of generating 3- and 4-wise independent pseudo-random numbers that are fast range-summable (i.e., they can be summed in sublinear time). Our specific contributions are: (a) we provide a thorough comparison of the various pseudo-random number generating schemes; (b) we study both theoretically and empirically the fast range-summation property of 3- and 4-wise independent generating schemes; (c) we provide algorithms for the fast range-summation of two 3-wise independent schemes, BCH and extended Hamming; and (d) we show convincing theoretical and empirical evidence that the extended Hamming scheme performs as well as any 4-wise independent scheme for estimating the size of join of two relations using AMS sketches, even though it is only 3-wise independent. We use this scheme to generate estimators that significantly outperform state-of-the-art solutions for two problems, namely, size of spatial joins and selectivity estimation .

Exact computation for aggregate queries usually requires large amounts of memory - constrained in data-streaming - or communication - constrained in distributed computation - and large processing times. In this situation, approximation techniques with provable guarantees, like sketches, are the only viable solution. The performance of sketches crucially depends on the ability to efficiently generate particular pseudo-random numbers. In this paper we investigate both theoretically and empirically the problem of generating k-wise independent pseudo-random numbers and, in particular, that of generating 3 and 4-wise independent pseudo-random numbers that are fast range-summable (i.e., they can be summed up in sub-linear time). Our specific contributions are: (a) we provide an empirical comparison of the various pseudo-random number generating schemes, (b) we study both theoretically and empirically the fast range-summation practicality for the 3 and 4-wise independent generating schemes and we provide efficient implementations for the 3-wise independent schemes, (c) we show convincing theoretical and empirical evidence that the extended Hamming scheme performs as well as any 4-wise independent scheme for estimating the size of join using AMS-sketches, even though it is only 3-wise independent. We use this generating scheme to produce estimators that significantly out-perform the state-of-the-art solutions for two problems - size of spatial joins and selectivity estimation.