DIMA and DFKI Researchers@VLDB 2018 [1]

The DIMA researchers mentioned some program topics as their personal highlights:

• Blockchain Tutorial by UC-Santa Barbara, a tutorial which pointed out chances, risks, pro’s, and con’s of this technology. A tutorial made by people who has no stake in any blockchain technology.
• Time Series Tutorial
• ADMS Keynotes: Quantum Computing and SIMD
• Industry Talk: Machine Learning for Online Advertising (Google)
• Amazon Time Series Tutorial (https://lovvge.github.io/Forecasting-Tutorial-VLDB-2018 [4])

... and gave a list of recommended reads:

• Moment-based Quantile Sketches for Efficient High Cardinality Aggregate Queries
• Improved Selectivity Estimation by Combining Knowledge from Sampling and Synopses
• Join Query Optimization Techniques for Complex Event Processing Applications (Ilya Kolchinsky and Assaf Schuster; Technion, Israel Institute of Technology)
• On Optimizing Operator Fusion Plans for Large-scale Machine Learning in SystemML, Boehm, M., Reinwald, B., Hutchison, D., Sen, P., Evfimievski, A. V., & Pansare, N.
• Filter Before you Parse: Faster Analytics on Raw Data with Sparser, Palkar, S., Abuzaid, F., Bailis, P., & Zaharia, M.

» » BlockJoin: Efficient Matrix Partitioning Through Joins, Andreas Kunft, Asterios Katsifodimos, Sebastian Schelter, Tilmann Rabl, Volker Markl.

Abstract: Linear algebra operations are at the core of many Machine Learning (ML) programs. At the same time, a considerable amount of the effort for solving data analytics problems is spent in data preparation. As a result, end-to-end ML pipelines often consist of (i) relational operators used for joining the input data, (ii) user defined functions used for feature extraction and vectorization, and (iii) linear algebra operators used for model training and cross-validation. Often, these pipelines need to scale out to large datasets. As a result, pipelines combining both kinds of operators require rather expensive changes to the physical representation, in particular re-partitioning steps. We present BlockJoin, a distributed join algorithm which directly produces block-partitioned results. To minimize shuffling costs, BlockJoin applies database techniques known from columnar processing, such as index-joins and late materialization, in the context of parallel dataflow engines.

Download Poster [8] | Link to the paper [9]

» » Estimating Join Selectivities using Bandwidth-Optimized Kernel Density Models, Martin Kiefer, Max Heiml, Sebastian Breß, Volker Markl.

Abstract: Accurately predicting the cardinality of intermediate plan operations is an essential part of any modern relational query optimizer. The accuracy of said estimates has a strong and direct impact on the quality of the generated plans, and incorrect estimates can have a negative impact on query performance. One of the biggest challenges in this field is to predict the result size of join operations. In this paper, we extend a bandwidth-optimized KDE models to estimate the result size of single and multiple joins. In particular, we propose two approaches: (1) Building a KDE model from a sample drawn from the join result. (2) Efficiently combining the information from base table KDE models.

Link to the paper [10]

» » Generating Custom Code for Efficient Query Code Execution on Heterogeneuos Processors, Sebastian Breß, Bastian Köcher, Henning Funke, Steffen Zeuch, Tilmann Rabl, Volker Markl.

Abstract: Accurately predicting the cardinality of intermediate plan operations is an essential part of any modern relational query optimizer. The accuracy of said estimates has a strong and direct impact on the quality of the generated plans, and incorrect estimates can have a negative impact on query performance. One of the biggest challenges in this field is to predict the result size of join operations. Kernel Density Estimation (KDE) is a statistical method to estimate multivariate probability distributions from a data sample. Previously, we introduced
a modern, self-tuning selectivity estimator for range scans based on KDE that outperforms state-of-the-art multidimensional histograms and is efficient to evaluate on graphics cards. In this paper, we extend these bandwidth-optimized KDE models to estimate the result size of single and multiple joins. In particular, we propose two approaches: (1) Building a KDE model from a sample drawn from the join result. (2) Efficiently combining the information from base table KDE models. We evaluated our KDE-based join estimators on a variety of synthetic and real-world datasets, demonstrating that they are superior to state-of-the art join estimators based on sketching or sampling.

Link to the paper [11]