Data sources, such as social media, mobile apps and IoT sensors, generate billions of records each day. Keeping up with this influx of data while providing useful analytics to the users is a major challenge for today's data-intensive systems. A popular solution that allows such systems to handle rapidly incoming data is to rely on log-structured merge (LSM) storage models. LSM-based systems provide a tunable trade-off between ingesting vast amounts of data at a high rate and running efficient analytical queries on top of that data. For queries, it is well-known that the query processing performance largely depends on the ability to generate efficient execution plans. Previous research showed that OLAP query workloads rely on having small, yet precise, statistical summaries of the underlying data, which can drive the cost-based query optimization. In this paper we address the problem of computing data statistics for workloads with rapid data ingestion and propose a lightweight statistics-collection framework that exploits the properties of LSM storage. Our approach is designed to piggyback on the events (flush and merge) of the LSM lifecycle. This allows us to easily create an initial statistics and then keep them in sync with rapidly changing data while minimizing the overhead to the existing system. We have implemented and adapted well-known algorithms to produce various types of statistical synopses, including equi-width histograms, equi-height histograms, and wavelets. We performed an in-depth empirical evaluation that considers both the cardinality estimation accuracy and runtime overheads of collecting and using statistics. The experiments were conducted by prototyping our approach on top of Apache AsterixDB, an open source Big Data management system that has an entirely LSM-based storage backend.
{"title":"Lightweight Cardinality Estimation in LSM-based Systems","authors":"Ildar Absalyamov, M. Carey, V. Tsotras","doi":"10.1145/3183713.3183761","DOIUrl":"https://doi.org/10.1145/3183713.3183761","url":null,"abstract":"Data sources, such as social media, mobile apps and IoT sensors, generate billions of records each day. Keeping up with this influx of data while providing useful analytics to the users is a major challenge for today's data-intensive systems. A popular solution that allows such systems to handle rapidly incoming data is to rely on log-structured merge (LSM) storage models. LSM-based systems provide a tunable trade-off between ingesting vast amounts of data at a high rate and running efficient analytical queries on top of that data. For queries, it is well-known that the query processing performance largely depends on the ability to generate efficient execution plans. Previous research showed that OLAP query workloads rely on having small, yet precise, statistical summaries of the underlying data, which can drive the cost-based query optimization. In this paper we address the problem of computing data statistics for workloads with rapid data ingestion and propose a lightweight statistics-collection framework that exploits the properties of LSM storage. Our approach is designed to piggyback on the events (flush and merge) of the LSM lifecycle. This allows us to easily create an initial statistics and then keep them in sync with rapidly changing data while minimizing the overhead to the existing system. We have implemented and adapted well-known algorithms to produce various types of statistical synopses, including equi-width histograms, equi-height histograms, and wavelets. We performed an in-depth empirical evaluation that considers both the cardinality estimation accuracy and runtime overheads of collecting and using statistics. The experiments were conducted by prototyping our approach on top of Apache AsterixDB, an open source Big Data management system that has an entirely LSM-based storage backend.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"3 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89775183","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Membership testing is the problem of testing whether an element is in a set of elements. Performing the test exactly is expensive space-wise, requiring the storage of all elements in a set. In many applications, an approximate testing that can be done quickly using small space is often desired. Bloom filter (BF) was designed and has witnessed great success across numerous application domains. But there is no compact structure that supports set membership testing for temporal queries, e.g., has person A visited a web server between 9:30am and 9:40am? And has the same person visited the web server again between 9:45am and 9:50am? It is possible to support such "temporal membership testing" using a BF, but we will show that this is fairly expensive. To that end, this paper designs persistent bloom filter (PBF), a novel data structure for temporal membership testing with compact space.
{"title":"Persistent Bloom Filter: Membership Testing for the Entire History","authors":"Yanqing Peng, Jinwei Guo, Feifei Li, Weining Qian, Aoying Zhou","doi":"10.1145/3183713.3183737","DOIUrl":"https://doi.org/10.1145/3183713.3183737","url":null,"abstract":"Membership testing is the problem of testing whether an element is in a set of elements. Performing the test exactly is expensive space-wise, requiring the storage of all elements in a set. In many applications, an approximate testing that can be done quickly using small space is often desired. Bloom filter (BF) was designed and has witnessed great success across numerous application domains. But there is no compact structure that supports set membership testing for temporal queries, e.g., has person A visited a web server between 9:30am and 9:40am? And has the same person visited the web server again between 9:45am and 9:50am? It is possible to support such \"temporal membership testing\" using a BF, but we will show that this is fairly expensive. To that end, this paper designs persistent bloom filter (PBF), a novel data structure for temporal membership testing with compact space.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"25 3 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88286829","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Adam Dziedzic, Jingjing Wang, Sudipto Das, Bolin Ding, Vivek R. Narasayya, M. Syamala
Commercial DBMSs, such as Microsoft SQL Server, cater to diverse workloads including transaction processing, decision support, and operational analytics. They also support variety in physical design structures such as B+ tree and columnstore. The benefits of B+ tree for OLTP workloads and columnstore for decision support workloads are well-understood. However, the importance of hybrid physical designs, consisting of both columnstore and B+ tree indexes on the same database, is not well-studied --- a focus of this paper. We first quantify the trade-offs using carefully-crafted micro-benchmarks. This micro-benchmarking indicates that hybrid physical designs can result in orders of magnitude better performance depending on the workload. For complex real-world applications, choosing an appropriate combination of columnstore and B+ tree indexes for a database workload is challenging. We extend the Database Engine Tuning Advisor for Microsoft SQL Server to recommend a suitable combination of B+ tree and columnstore indexes for a given workload. Through extensive experiments using industry-standard benchmarks and several real-world customer workloads, we quantify how a physical design tool capable of recommending hybrid physical designs can result in orders of magnitude better execution costs compared to approaches that rely either on columnstore-only or B+ tree-only designs.
{"title":"Columnstore and B+ tree - Are Hybrid Physical Designs Important?","authors":"Adam Dziedzic, Jingjing Wang, Sudipto Das, Bolin Ding, Vivek R. Narasayya, M. Syamala","doi":"10.1145/3183713.3190660","DOIUrl":"https://doi.org/10.1145/3183713.3190660","url":null,"abstract":"Commercial DBMSs, such as Microsoft SQL Server, cater to diverse workloads including transaction processing, decision support, and operational analytics. They also support variety in physical design structures such as B+ tree and columnstore. The benefits of B+ tree for OLTP workloads and columnstore for decision support workloads are well-understood. However, the importance of hybrid physical designs, consisting of both columnstore and B+ tree indexes on the same database, is not well-studied --- a focus of this paper. We first quantify the trade-offs using carefully-crafted micro-benchmarks. This micro-benchmarking indicates that hybrid physical designs can result in orders of magnitude better performance depending on the workload. For complex real-world applications, choosing an appropriate combination of columnstore and B+ tree indexes for a database workload is challenging. We extend the Database Engine Tuning Advisor for Microsoft SQL Server to recommend a suitable combination of B+ tree and columnstore indexes for a given workload. Through extensive experiments using industry-standard benchmarks and several real-world customer workloads, we quantify how a physical design tool capable of recommending hybrid physical designs can result in orders of magnitude better execution costs compared to approaches that rely either on columnstore-only or B+ tree-only designs.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"204 2","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72562732","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
{"title":"Session details: Keynote 2","authors":"Xinyue Dong","doi":"10.1145/3258012","DOIUrl":"https://doi.org/10.1145/3258012","url":null,"abstract":"","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"18 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75739739","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Michele Linardi, Yan Zhu, Themis Palpanas, Eamonn J. Keogh
Data series motif discovery represents one of the most useful primitives for data series mining, with applications to many domains, such as robotics, entomology, seismology, medicine, and climatology, and others. The state-of-the-art motif discovery tools still require the user to provide the motif length. Yet, in several cases, the choice of motif length is critical for their detection. Unfortunately, the obvious brute-force solution, which tests all lengths within a given range, is computationally untenable, and does not provide any support for ranking motifs at different resolutions (i.e., lengths). We demonstrate VALMOD, our scalable motif discovery algorithm that efficiently finds all motifs in a given range of lengths, and outputs a length-invariant ranking of motifs. Furthermore, we support the analysis process by means of a newly proposed meta-data structure that helps the user to select the most promising pattern length. This demo aims at illustrating in detail the steps of the proposed approach, showcasing how our algorithm and corresponding graphical insights enable users to efficiently identify the correct motifs.
{"title":"VALMOD: A Suite for Easy and Exact Detection of Variable Length Motifs in Data Series","authors":"Michele Linardi, Yan Zhu, Themis Palpanas, Eamonn J. Keogh","doi":"10.1145/3183713.3193556","DOIUrl":"https://doi.org/10.1145/3183713.3193556","url":null,"abstract":"Data series motif discovery represents one of the most useful primitives for data series mining, with applications to many domains, such as robotics, entomology, seismology, medicine, and climatology, and others. The state-of-the-art motif discovery tools still require the user to provide the motif length. Yet, in several cases, the choice of motif length is critical for their detection. Unfortunately, the obvious brute-force solution, which tests all lengths within a given range, is computationally untenable, and does not provide any support for ranking motifs at different resolutions (i.e., lengths). We demonstrate VALMOD, our scalable motif discovery algorithm that efficiently finds all motifs in a given range of lengths, and outputs a length-invariant ranking of motifs. Furthermore, we support the analysis process by means of a newly proposed meta-data structure that helps the user to select the most promising pattern length. This demo aims at illustrating in detail the steps of the proposed approach, showcasing how our algorithm and corresponding graphical insights enable users to efficiently identify the correct motifs.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"201 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"76998094","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To address the challenge of explosive big data, distributed machine learning (ML) has drawn the interests of many researchers. Since many distributed ML algorithms trained by stochastic gradient descent (SGD) involve communicating gradients through the network, it is important to compress the transferred gradient. A category of low-precision algorithms can significantly reduce the size of gradients, at the expense of some precision loss. However, existing low-precision methods are not suitable for many cases where the gradients are sparse and nonuniformly distributed. In this paper, we study is there a compression method that can efficiently handle a sparse and nonuniform gradient consisting of key-value pairs? Our first contribution is a sketch based method that compresses the gradient values. Sketch is a class of algorithms using a probabilistic data structure to approximate the distribution of input data. We design a quantile-bucket quantification method that uses a quantile sketch to sort gradient values into buckets and encodes them with the bucket indexes. To further compress the bucket indexes, our second contribution is a sketch algorithm, namely MinMaxSketch. MinMaxSketch builds a set of hash tables and solves hash collisions with a MinMax strategy. The third contribution of this paper is a delta-binary encoding method that calculates the increment of the gradient keys and stores them with fewer bytes. We also theoretically discuss the correctness and the error bound of three proposed methods. To the best of our knowledge, this is the first effort combining data sketch with ML. We implement a prototype system in a real cluster of our industrial partner Tencent Inc., and show that our method is up to 10X faster than existing methods.
{"title":"SketchML: Accelerating Distributed Machine Learning with Data Sketches","authors":"Jiawei Jiang, Fangcheng Fu, Tong Yang, B. Cui","doi":"10.1145/3183713.3196894","DOIUrl":"https://doi.org/10.1145/3183713.3196894","url":null,"abstract":"To address the challenge of explosive big data, distributed machine learning (ML) has drawn the interests of many researchers. Since many distributed ML algorithms trained by stochastic gradient descent (SGD) involve communicating gradients through the network, it is important to compress the transferred gradient. A category of low-precision algorithms can significantly reduce the size of gradients, at the expense of some precision loss. However, existing low-precision methods are not suitable for many cases where the gradients are sparse and nonuniformly distributed. In this paper, we study is there a compression method that can efficiently handle a sparse and nonuniform gradient consisting of key-value pairs? Our first contribution is a sketch based method that compresses the gradient values. Sketch is a class of algorithms using a probabilistic data structure to approximate the distribution of input data. We design a quantile-bucket quantification method that uses a quantile sketch to sort gradient values into buckets and encodes them with the bucket indexes. To further compress the bucket indexes, our second contribution is a sketch algorithm, namely MinMaxSketch. MinMaxSketch builds a set of hash tables and solves hash collisions with a MinMax strategy. The third contribution of this paper is a delta-binary encoding method that calculates the increment of the gradient keys and stores them with fewer bytes. We also theoretically discuss the correctness and the error bound of three proposed methods. To the best of our knowledge, this is the first effort combining data sketch with ML. We implement a prototype system in a real cluster of our industrial partner Tencent Inc., and show that our method is up to 10X faster than existing methods.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"23 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"80079774","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Rebecca Taft, Nosayba El-Sayed, M. Serafini, Yu Lu, Ashraf Aboulnaga, M. Stonebraker, Ricardo Mayerhofer, Francisco Jose Andrade
OLTP database systems are a critical part of the operation of many enterprises. Such systems are often configured statically with sufficient capacity for peak load. For many OLTP applications, however, the maximum load is an order of magnitude larger than the minimum, and load varies in a repeating daily pattern. It is thus prudent to allocate computing resources dynamically to match demand. One can allocate resources reactively after a load increase is detected, but this places additional burden on the already-overloaded system to reconfigure. A predictive allocation, in advance of load increases, is clearly preferable. We present P-Store, the first elastic OLTP DBMS to use prediction, and apply it to the workload of B2W Digital (B2W), a large online retailer. Our study shows that P-Store outperforms a reactive system on B2W's workload by causing 72% fewer latency violations, and achieves performance comparable to static allocation for peak demand while using 50% fewer servers.
OLTP数据库系统是许多企业运营的关键部分。这样的系统通常静态配置，具有足够的峰值负载容量。但是，对于许多OLTP应用程序，最大负载比最小负载大一个数量级，并且负载以每天重复的模式变化。因此，动态分配计算资源以匹配需求是明智的。可以在检测到负载增加后响应性地分配资源，但这会给已经过载的系统带来额外的负担，需要重新配置。在负载增加之前进行预测分配显然是可取的。我们提出了P-Store，这是第一个使用预测的弹性OLTP DBMS，并将其应用于大型在线零售商B2W Digital (B2W)的工作负载。我们的研究表明，在B2W的工作负载上，P-Store的性能优于响应式系统，因为它导致的延迟违规减少了72%，并且在使用50%的服务器时实现了与峰值需求静态分配相当的性能。
{"title":"P-Store: An Elastic Database System with Predictive Provisioning","authors":"Rebecca Taft, Nosayba El-Sayed, M. Serafini, Yu Lu, Ashraf Aboulnaga, M. Stonebraker, Ricardo Mayerhofer, Francisco Jose Andrade","doi":"10.1145/3183713.3190650","DOIUrl":"https://doi.org/10.1145/3183713.3190650","url":null,"abstract":"OLTP database systems are a critical part of the operation of many enterprises. Such systems are often configured statically with sufficient capacity for peak load. For many OLTP applications, however, the maximum load is an order of magnitude larger than the minimum, and load varies in a repeating daily pattern. It is thus prudent to allocate computing resources dynamically to match demand. One can allocate resources reactively after a load increase is detected, but this places additional burden on the already-overloaded system to reconfigure. A predictive allocation, in advance of load increases, is clearly preferable. We present P-Store, the first elastic OLTP DBMS to use prediction, and apply it to the workload of B2W Digital (B2W), a large online retailer. Our study shows that P-Store outperforms a reactive system on B2W's workload by causing 72% fewer latency violations, and achieves performance comparable to static allocation for peak demand while using 50% fewer servers.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"16 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81501632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Jinglin Peng, Dongxiang Zhang, Jiannan Wang, J. Pei
Interactive analytics requires database systems to be able to answer aggregation queries within interactive response times. As the amount of data is continuously growing at an unprecedented rate, this is becoming increasingly challenging. In the past, the database community has proposed two separate ideas, sampling-based approximate query processing (AQP) and aggregate precomputation (AggPre) such as data cubes, to address this challenge. In this paper, we argue for the need to connect these two separate ideas for interactive analytics. We propose AQP++, a novel framework to enable the connection. The framework can leverage both a sample as well as a precomputed aggregate to answer user queries. We discuss the advantages of having such a unified framework and identify new challenges to fulfill this vision. We conduct an in-depth study of these challenges for range queries and explore both optimal and heuristic solutions to address them. Our experiments using two public benchmarks and one real-world dataset show that AQP++ achieves a more flexible and better trade-off among preprocessing cost, query response time, and answer quality than AQP or AggPre.
{"title":"AQP++: Connecting Approximate Query Processing With Aggregate Precomputation for Interactive Analytics","authors":"Jinglin Peng, Dongxiang Zhang, Jiannan Wang, J. Pei","doi":"10.1145/3183713.3183747","DOIUrl":"https://doi.org/10.1145/3183713.3183747","url":null,"abstract":"Interactive analytics requires database systems to be able to answer aggregation queries within interactive response times. As the amount of data is continuously growing at an unprecedented rate, this is becoming increasingly challenging. In the past, the database community has proposed two separate ideas, sampling-based approximate query processing (AQP) and aggregate precomputation (AggPre) such as data cubes, to address this challenge. In this paper, we argue for the need to connect these two separate ideas for interactive analytics. We propose AQP++, a novel framework to enable the connection. The framework can leverage both a sample as well as a precomputed aggregate to answer user queries. We discuss the advantages of having such a unified framework and identify new challenges to fulfill this vision. We conduct an in-depth study of these challenges for range queries and explore both optimal and heuristic solutions to address them. Our experiments using two public benchmarks and one real-world dataset show that AQP++ achieves a more flexible and better trade-off among preprocessing cost, query response time, and answer quality than AQP or AggPre.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"121 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90561826","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Mustafa Korkmaz, M. Karsten, K. Salem, S. Salihoglu
Natural short term fluctuations in the load of transactional data systems present an opportunity for power savings. For example, a system handling 1000 requests per second on average can expect more than 1000 requests in some seconds, fewer in others. By quickly adjusting processing capacity to match such fluctuations, power consumption can be reduced. Many systems do this already, using dynamic voltage and frequency scaling (DVFS) to reduce processor performance and power consumption when the load is low. DVFS is typically controlled by frequency governors in the operating system, or by the processor itself. In this paper, we show that transactional database systems can manage DVFS more effectively than the underlying operating system. This is because the database system has more information about the workload, and more control over that workload, than is available to the operating system. We present a technique called POLARIS for reducing the power consumption of transactional database systems. POLARIS directly manages processor DVFS and controls database transaction scheduling. Its goal is to minimize power consumption while ensuring the transactions are completed within a specified latency target. POLARIS is workload-aware, and can accommodate concurrent workloads with different characteristics and latency budgets. We show that POLARIS can simultaneously reduce power consumption and reduce missed latency targets, relative to operating-system-based DVFS governors.
{"title":"Workload-Aware CPU Performance Scaling for Transactional Database Systems","authors":"Mustafa Korkmaz, M. Karsten, K. Salem, S. Salihoglu","doi":"10.1145/3183713.3196901","DOIUrl":"https://doi.org/10.1145/3183713.3196901","url":null,"abstract":"Natural short term fluctuations in the load of transactional data systems present an opportunity for power savings. For example, a system handling 1000 requests per second on average can expect more than 1000 requests in some seconds, fewer in others. By quickly adjusting processing capacity to match such fluctuations, power consumption can be reduced. Many systems do this already, using dynamic voltage and frequency scaling (DVFS) to reduce processor performance and power consumption when the load is low. DVFS is typically controlled by frequency governors in the operating system, or by the processor itself. In this paper, we show that transactional database systems can manage DVFS more effectively than the underlying operating system. This is because the database system has more information about the workload, and more control over that workload, than is available to the operating system. We present a technique called POLARIS for reducing the power consumption of transactional database systems. POLARIS directly manages processor DVFS and controls database transaction scheduling. Its goal is to minimize power consumption while ensuring the transactions are completed within a specified latency target. POLARIS is workload-aware, and can accommodate concurrent workloads with different characteristics and latency budgets. We show that POLARIS can simultaneously reduce power consumption and reduce missed latency targets, relative to operating-system-based DVFS governors.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"6 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90579089","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Sidharth Mudgal, Han Li, Theodoros Rekatsinas, A. Doan, Youngchoon Park, Ganesh Krishnan, Rohit Deep, Esteban Arcaute, V. Raghavendra
Entity matching (EM) finds data instances that refer to the same real-world entity. In this paper we examine applying deep learning (DL) to EM, to understand DL's benefits and limitations. We review many DL solutions that have been developed for related matching tasks in text processing (e.g., entity linking, textual entailment, etc.). We categorize these solutions and define a space of DL solutions for EM, as embodied by four solutions with varying representational power: SIF, RNN, Attention, and Hybrid. Next, we investigate the types of EM problems for which DL can be helpful. We consider three such problem types, which match structured data instances, textual instances, and dirty instances, respectively. We empirically compare the above four DL solutions with Magellan, a state-of-the-art learning-based EM solution. The results show that DL does not outperform current solutions on structured EM, but it can significantly outperform them on textual and dirty EM. For practitioners, this suggests that they should seriously consider using DL for textual and dirty EM problems. Finally, we analyze DL's performance and discuss future research directions.
{"title":"Deep Learning for Entity Matching: A Design Space Exploration","authors":"Sidharth Mudgal, Han Li, Theodoros Rekatsinas, A. Doan, Youngchoon Park, Ganesh Krishnan, Rohit Deep, Esteban Arcaute, V. Raghavendra","doi":"10.1145/3183713.3196926","DOIUrl":"https://doi.org/10.1145/3183713.3196926","url":null,"abstract":"Entity matching (EM) finds data instances that refer to the same real-world entity. In this paper we examine applying deep learning (DL) to EM, to understand DL's benefits and limitations. We review many DL solutions that have been developed for related matching tasks in text processing (e.g., entity linking, textual entailment, etc.). We categorize these solutions and define a space of DL solutions for EM, as embodied by four solutions with varying representational power: SIF, RNN, Attention, and Hybrid. Next, we investigate the types of EM problems for which DL can be helpful. We consider three such problem types, which match structured data instances, textual instances, and dirty instances, respectively. We empirically compare the above four DL solutions with Magellan, a state-of-the-art learning-based EM solution. The results show that DL does not outperform current solutions on structured EM, but it can significantly outperform them on textual and dirty EM. For practitioners, this suggests that they should seriously consider using DL for textual and dirty EM problems. Finally, we analyze DL's performance and discuss future research directions.","PeriodicalId":20430,"journal":{"name":"Proceedings of the 2018 International Conference on Management of Data","volume":"22 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2018-05-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83494672","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：