T. Kissinger, B. Schlegel, Dirk Habich, Wolfgang Lehner
Modern database systems have to process huge amounts of data and should provide results with low latency at the same time. To achieve this, data is nowadays typically hold completely in main memory, to benefit of its high bandwidth and low access latency that could never be reached with disks. Current in-memory databases are usually column-stores that exchange columns or vectors between operators and suffer from a high tuple reconstruction overhead. In this demonstration proposal, we present DexterDB, which implements our novel prefix tree-based processing model that makes indexes the first-class citizen of the database system. The core idea is that each operator takes a set of indexes as input and builds a new index as output that is indexed on the attribute requested by the successive operator. With that, we are able to build composed operators, like the multi-way-select-join-group. Such operators speed up the processing of complex OLAP queries so that DexterDB outperforms state-of-the-art in-memory databases. Our demonstration focuses on the different optimization options for such query plans. Hence, we built an interactive GUI that connects to a DexterDB instance and allows the manipulation of query optimization parameters. The generated query plans and important execution statistics are visualized to help the visitor to understand our processing model.
{"title":"Query processing on prefix trees live","authors":"T. Kissinger, B. Schlegel, Dirk Habich, Wolfgang Lehner","doi":"10.1145/2463676.2463682","DOIUrl":"https://doi.org/10.1145/2463676.2463682","url":null,"abstract":"Modern database systems have to process huge amounts of data and should provide results with low latency at the same time. To achieve this, data is nowadays typically hold completely in main memory, to benefit of its high bandwidth and low access latency that could never be reached with disks. Current in-memory databases are usually column-stores that exchange columns or vectors between operators and suffer from a high tuple reconstruction overhead. In this demonstration proposal, we present DexterDB, which implements our novel prefix tree-based processing model that makes indexes the first-class citizen of the database system. The core idea is that each operator takes a set of indexes as input and builds a new index as output that is indexed on the attribute requested by the successive operator. With that, we are able to build composed operators, like the multi-way-select-join-group. Such operators speed up the processing of complex OLAP queries so that DexterDB outperforms state-of-the-art in-memory databases. Our demonstration focuses on the different optimization options for such query plans. Hence, we built an interactive GUI that connects to a DexterDB instance and allows the manipulation of query optimization parameters. The generated query plans and important execution statistics are visualized to help the visitor to understand our processing model.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85505012","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}
Meihui Zhang, Hazem Elmeleegy, Cecilia M. Procopiuc, D. Srivastava
We study the following problem: Given a database D with schema G and an output table Out, compute a join query Q that generates OUT from D. A simpler variant allows Q to return a superset of Out. This problem has numerous applications, both by itself, and as a building block for other problems. Related prior work imposes conditions on the structure of Q which are not always consistent with the application, but simplify computation. We discuss several natural SQL queries that do not satisfy these conditions and cannot be discovered by prior work. In this paper, we propose an efficient algorithm that discovers queries with arbitrary join graphs. A crucial insight is that any graph can be characterized by the combination of a simple structure, called a star, and a series of merge steps over the star. The merge steps define a lattice over graphs derived from the same star. This allows us to explore the set of candidate solutions in a principled way and quickly prune out a large number of infeasible graphs. We also design several optimizations that significantly reduce the running time. Finally, we conduct an extensive experimental study over a benchmark database and show that our approach is scalable and accurately discovers complex join queries.
{"title":"Reverse engineering complex join queries","authors":"Meihui Zhang, Hazem Elmeleegy, Cecilia M. Procopiuc, D. Srivastava","doi":"10.1145/2463676.2465320","DOIUrl":"https://doi.org/10.1145/2463676.2465320","url":null,"abstract":"We study the following problem: Given a database D with schema G and an output table Out, compute a join query Q that generates OUT from D. A simpler variant allows Q to return a superset of Out. This problem has numerous applications, both by itself, and as a building block for other problems. Related prior work imposes conditions on the structure of Q which are not always consistent with the application, but simplify computation. We discuss several natural SQL queries that do not satisfy these conditions and cannot be discovered by prior work.\u0000 In this paper, we propose an efficient algorithm that discovers queries with arbitrary join graphs. A crucial insight is that any graph can be characterized by the combination of a simple structure, called a star, and a series of merge steps over the star. The merge steps define a lattice over graphs derived from the same star. This allows us to explore the set of candidate solutions in a principled way and quickly prune out a large number of infeasible graphs. We also design several optimizations that significantly reduce the running time. Finally, we conduct an extensive experimental study over a benchmark database and show that our approach is scalable and accurately discovers complex join queries.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82784276","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}
Patricia C. Arocena, Boris Glavic, Renée J. Miller
The creation of values to represent incomplete information, often referred to as value invention, is central in data exchange. Within schema mappings, Skolem functions have long been used for value invention as they permit a precise representation of missing information. Recent work on a powerful mapping language called second-order tuple generating dependencies (SO tgds), has drawn attention to the fact that the use of arbitrary Skolem functions can have negative computational and programmatic properties in data exchange. In this paper, we present two techniques for understanding when the Skolem functions needed to represent the correct semantics of incomplete information are computationally well-behaved. Specifically, we consider when the Skolem functions in second-order (SO) mappings have a first-order (FO) semantics and are therefore programmatically and computationally more desirable for use in practice. Our first technique, linearization, significantly extends the Nash, Bernstein and Melnik unskolemization algorithm, by understanding when the sets of arguments of the Skolem functions in a mapping are related by set inclusion. We show that such a linear relationship leads to mappings that have FO semantics and are expressible in popular mapping languages including source-to-target tgds and nested tgds. Our second technique uses source semantics, specifically functional dependencies (including keys), to transform SO mappings into equivalent FO mappings. We show that our algorithms are applicable to a strictly larger class of mappings than previous approaches, but more importantly we present an extensive experimental evaluation that quantifies this difference (about 78% improvement) over an extensive schema mapping benchmark and illustrates the applicability of our results on real mappings.
{"title":"Value invention in data exchange","authors":"Patricia C. Arocena, Boris Glavic, Renée J. Miller","doi":"10.1145/2463676.2465311","DOIUrl":"https://doi.org/10.1145/2463676.2465311","url":null,"abstract":"The creation of values to represent incomplete information, often referred to as value invention, is central in data exchange. Within schema mappings, Skolem functions have long been used for value invention as they permit a precise representation of missing information. Recent work on a powerful mapping language called second-order tuple generating dependencies (SO tgds), has drawn attention to the fact that the use of arbitrary Skolem functions can have negative computational and programmatic properties in data exchange. In this paper, we present two techniques for understanding when the Skolem functions needed to represent the correct semantics of incomplete information are computationally well-behaved. Specifically, we consider when the Skolem functions in second-order (SO) mappings have a first-order (FO) semantics and are therefore programmatically and computationally more desirable for use in practice. Our first technique, linearization, significantly extends the Nash, Bernstein and Melnik unskolemization algorithm, by understanding when the sets of arguments of the Skolem functions in a mapping are related by set inclusion. We show that such a linear relationship leads to mappings that have FO semantics and are expressible in popular mapping languages including source-to-target tgds and nested tgds. Our second technique uses source semantics, specifically functional dependencies (including keys), to transform SO mappings into equivalent FO mappings. We show that our algorithms are applicable to a strictly larger class of mappings than previous approaches, but more importantly we present an extensive experimental evaluation that quantifies this difference (about 78% improvement) over an extensive schema mapping benchmark and illustrates the applicability of our results on real mappings.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1145/2463676.2465311","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72435709","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}
Jinyang Gao, Xuan Liu, B. Ooi, Haixun Wang, Gang Chen
Crowdsourcing has created a variety of opportunities for many challenging problems by leveraging human intelligence. For example, applications such as image tagging, natural language processing, and semantic-based information retrieval can exploit crowd-based human computation to supplement existing computational algorithms. Naturally, human workers in crowdsourcing solve problems based on their knowledge, experience, and perception. It is therefore not clear which problems can be better solved by crowdsourcing than solving solely using traditional machine-based methods. Therefore, a cost sensitive quantitative analysis method is needed. In this paper, we design and implement a cost sensitive method for crowdsourcing. We online estimate the profit of the crowdsourcing job so that those questions with no future profit from crowdsourcing can be terminated. Two models are proposed to estimate the profit of crowdsourcing job, namely the linear value model and the generalized non-linear model. Using these models, the expected profit of obtaining new answers for a specific question is computed based on the answers already received. A question is terminated in real time if the marginal expected profit of obtaining more answers is not positive. We extends the method to publish a batch of questions in a HIT. We evaluate the effectiveness of our proposed method using two real world jobs on AMT. The experimental results show that our proposed method outperforms all the state-of-art methods.
{"title":"An online cost sensitive decision-making method in crowdsourcing systems","authors":"Jinyang Gao, Xuan Liu, B. Ooi, Haixun Wang, Gang Chen","doi":"10.1145/2463676.2465307","DOIUrl":"https://doi.org/10.1145/2463676.2465307","url":null,"abstract":"Crowdsourcing has created a variety of opportunities for many challenging problems by leveraging human intelligence. For example, applications such as image tagging, natural language processing, and semantic-based information retrieval can exploit crowd-based human computation to supplement existing computational algorithms. Naturally, human workers in crowdsourcing solve problems based on their knowledge, experience, and perception. It is therefore not clear which problems can be better solved by crowdsourcing than solving solely using traditional machine-based methods. Therefore, a cost sensitive quantitative analysis method is needed.\u0000 In this paper, we design and implement a cost sensitive method for crowdsourcing. We online estimate the profit of the crowdsourcing job so that those questions with no future profit from crowdsourcing can be terminated. Two models are proposed to estimate the profit of crowdsourcing job, namely the linear value model and the generalized non-linear model. Using these models, the expected profit of obtaining new answers for a specific question is computed based on the answers already received. A question is terminated in real time if the marginal expected profit of obtaining more answers is not positive. We extends the method to publish a batch of questions in a HIT. We evaluate the effectiveness of our proposed method using two real world jobs on AMT. The experimental results show that our proposed method outperforms all the state-of-art methods.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74476154","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}
P. Upadhyaya, Nick R. Anderson, M. Balazinska, Bill Howe, R. Kaushik, Ravishankar Ramamurthy, Dan Suciu
In this demonstration, we show-case a database management system extended with a new type of component that we call a Data Use Manager (DUM). The DUM enables DBAs to attach policies to data loaded into the DBMS. It then monitors how users query the data, flags potential policy violations, recommends possible fixes, and supports offline analysis of user activities related to data policies. The demonstration uses real healthcare data.
在本演示中,我们展示了一个数据库管理系统,该系统扩展了一种称为数据使用管理器(Data Use Manager, DUM)的新型组件。DUM使dba能够将策略附加到加载到DBMS中的数据上。然后,它监视用户如何查询数据,标记潜在的策略违规,建议可能的修复,并支持与数据策略相关的用户活动的离线分析。该演示使用真实的医疗保健数据。
{"title":"The power of data use management in action","authors":"P. Upadhyaya, Nick R. Anderson, M. Balazinska, Bill Howe, R. Kaushik, Ravishankar Ramamurthy, Dan Suciu","doi":"10.1145/2463676.2465264","DOIUrl":"https://doi.org/10.1145/2463676.2465264","url":null,"abstract":"In this demonstration, we show-case a database management system extended with a new type of component that we call a Data Use Manager (DUM). The DUM enables DBAs to attach policies to data loaded into the DBMS. It then monitors how users query the data, flags potential policy violations, recommends possible fixes, and supports offline analysis of user activities related to data policies. The demonstration uses real healthcare data.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74543273","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}
Iris Miliaraki, K. Berberich, Rainer Gemulla, Spyros Zoupanos
Frequent sequence mining is one of the fundamental building blocks in data mining. While the problem has been extensively studied, few of the available techniques are sufficiently scalable to handle datasets with billions of sequences; such large-scale datasets arise, for instance, in text mining and session analysis. In this paper, we propose MG-FSM, a scalable algorithm for frequent sequence mining on MapReduce. MG-FSM can handle so-called "gap constraints", which can be used to limit the output to a controlled set of frequent sequences. At its heart, MG-FSM partitions the input database in a way that allows us to mine each partition independently using any existing frequent sequence mining algorithm. We introduce the notion of w-equivalency, which is a generalization of the notion of a "projected database" used by many frequent pattern mining algorithms. We also present a number of optimization techniques that minimize partition size, and therefore computational and communication costs, while still maintaining correctness. Our experimental study in the context of text mining suggests that MG-FSM is significantly more efficient and scalable than alternative approaches.
{"title":"Mind the gap: large-scale frequent sequence mining","authors":"Iris Miliaraki, K. Berberich, Rainer Gemulla, Spyros Zoupanos","doi":"10.1145/2463676.2465285","DOIUrl":"https://doi.org/10.1145/2463676.2465285","url":null,"abstract":"Frequent sequence mining is one of the fundamental building blocks in data mining. While the problem has been extensively studied, few of the available techniques are sufficiently scalable to handle datasets with billions of sequences; such large-scale datasets arise, for instance, in text mining and session analysis. In this paper, we propose MG-FSM, a scalable algorithm for frequent sequence mining on MapReduce. MG-FSM can handle so-called \"gap constraints\", which can be used to limit the output to a controlled set of frequent sequences. At its heart, MG-FSM partitions the input database in a way that allows us to mine each partition independently using any existing frequent sequence mining algorithm. We introduce the notion of w-equivalency, which is a generalization of the notion of a \"projected database\" used by many frequent pattern mining algorithms. We also present a number of optimization techniques that minimize partition size, and therefore computational and communication costs, while still maintaining correctness. Our experimental study in the context of text mining suggests that MG-FSM is significantly more efficient and scalable than alternative approaches.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75048058","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}
Tian Tan, Richard T. B. Ma, M. Winslett, Y. Yang, Yong Yu, Zhenjie Zhang
We propose Resa, a novel framework for robust, elastic and realtime stream processing in the cloud. In addition to traditional functionalities of streaming and cloud systems, Resa provides (i) a novel mechanism that handles dynamic additions and removals nodes in an operator, and (ii) a node re-assignment scheme that minimizes output latency using a queuing model. We have implemented Resa on top of Twitter Storm. Experiments using real data demonstrate the effectiveness and efficiency of Resa.
{"title":"Resa: realtime elastic streaming analytics in the cloud","authors":"Tian Tan, Richard T. B. Ma, M. Winslett, Y. Yang, Yong Yu, Zhenjie Zhang","doi":"10.1145/2463676.2465343","DOIUrl":"https://doi.org/10.1145/2463676.2465343","url":null,"abstract":"We propose Resa, a novel framework for robust, elastic and realtime stream processing in the cloud. In addition to traditional functionalities of streaming and cloud systems, Resa provides (i) a novel mechanism that handles dynamic additions and removals nodes in an operator, and (ii) a node re-assignment scheme that minimizes output latency using a queuing model. We have implemented Resa on top of Twitter Storm. Experiments using real data demonstrate the effectiveness and efficiency of Resa.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75057852","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}
Regular expression matching over sequences in real time is a crucial task in complex event processing on data streams. Given that such data sequences are often noisy and errors have temporal and spatial correlations, performing regular expression matching effectively and efficiently is a challenging task. Instead of the traditional approach of learning a distribution of the stream first and then processing queries, we propose a new approach that efficiently does the matching based on an error model. In particular, our algorithms are based on the realistic Markov chain error model, and report all matching paths to trace relevant basic events that trigger the matching. This is much more informative than a single matching path. We also devise algorithms to efficiently return only top-k matching paths, and to handle negations in an extended regular expression. Finally, we conduct a comprehensive experimental study to evaluate our algorithms using real datasets.
{"title":"ε-Matching: event processing over noisy sequences in real time","authors":"Zheng Li, Tingjian Ge, Cindy X. Chen","doi":"10.1145/2463676.2463715","DOIUrl":"https://doi.org/10.1145/2463676.2463715","url":null,"abstract":"Regular expression matching over sequences in real time is a crucial task in complex event processing on data streams. Given that such data sequences are often noisy and errors have temporal and spatial correlations, performing regular expression matching effectively and efficiently is a challenging task. Instead of the traditional approach of learning a distribution of the stream first and then processing queries, we propose a new approach that efficiently does the matching based on an error model. In particular, our algorithms are based on the realistic Markov chain error model, and report all matching paths to trace relevant basic events that trigger the matching. This is much more informative than a single matching path. We also devise algorithms to efficiently return only top-k matching paths, and to handle negations in an extended regular expression. Finally, we conduct a comprehensive experimental study to evaluate our algorithms using real datasets.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78161123","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}
C. Diaconu, Craig S. Freedman, Erik Ismert, P. Larson, Pravin Mittal, Ryan Stonecipher, Nitin Verma, M. Zwilling
Hekaton is a new database engine optimized for memory resident data and OLTP workloads. Hekaton is fully integrated into SQL Server; it is not a separate system. To take advantage of Hekaton, a user simply declares a table memory optimized. Hekaton tables are fully transactional and durable and accessed using T-SQL in the same way as regular SQL Server tables. A query can reference both Hekaton tables and regular tables and a transaction can update data in both types of tables. T-SQL stored procedures that reference only Hekaton tables can be compiled into machine code for further performance improvements. The engine is designed for high con-currency. To achieve this it uses only latch-free data structures and a new optimistic, multiversion concurrency control technique. This paper gives an overview of the design of the Hekaton engine and reports some experimental results.
{"title":"Hekaton: SQL server's memory-optimized OLTP engine","authors":"C. Diaconu, Craig S. Freedman, Erik Ismert, P. Larson, Pravin Mittal, Ryan Stonecipher, Nitin Verma, M. Zwilling","doi":"10.1145/2463676.2463710","DOIUrl":"https://doi.org/10.1145/2463676.2463710","url":null,"abstract":"Hekaton is a new database engine optimized for memory resident data and OLTP workloads. Hekaton is fully integrated into SQL Server; it is not a separate system. To take advantage of Hekaton, a user simply declares a table memory optimized. Hekaton tables are fully transactional and durable and accessed using T-SQL in the same way as regular SQL Server tables. A query can reference both Hekaton tables and regular tables and a transaction can update data in both types of tables. T-SQL stored procedures that reference only Hekaton tables can be compiled into machine code for further performance improvements. The engine is designed for high con-currency. To achieve this it uses only latch-free data structures and a new optimistic, multiversion concurrency control technique. This paper gives an overview of the design of the Hekaton engine and reports some experimental results.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"85985178","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}
The use of large-scale data mining and machine learning has proliferated through the adoption of technologies such as Hadoop, with its simple programming semantics and rich and active ecosystem. This paper presents LinkedIn's Hadoop-based analytics stack, which allows data scientists and machine learning researchers to extract insights and build product features from massive amounts of data. In particular, we present our solutions to the ``last mile'' issues in providing a rich developer ecosystem. This includes easy ingress from and egress to online systems, and managing workflows as production processes. A key characteristic of our solution is that these distributed system concerns are completely abstracted away from researchers. For example, deploying data back into the online system is simply a 1-line Pig command that a data scientist can add to the end of their script. We also present case studies on how this ecosystem is used to solve problems ranging from recommendations to news feed updates to email digesting to descriptive analytical dashboards for our members.
{"title":"The big data ecosystem at LinkedIn","authors":"Roshan Sumbaly, J. Kreps, Sam Shah","doi":"10.1145/2463676.2463707","DOIUrl":"https://doi.org/10.1145/2463676.2463707","url":null,"abstract":"The use of large-scale data mining and machine learning has proliferated through the adoption of technologies such as Hadoop, with its simple programming semantics and rich and active ecosystem. This paper presents LinkedIn's Hadoop-based analytics stack, which allows data scientists and machine learning researchers to extract insights and build product features from massive amounts of data. In particular, we present our solutions to the ``last mile'' issues in providing a rich developer ecosystem. This includes easy ingress from and egress to online systems, and managing workflows as production processes. A key characteristic of our solution is that these distributed system concerns are completely abstracted away from researchers. For example, deploying data back into the online system is simply a 1-line Pig command that a data scientist can add to the end of their script. We also present case studies on how this ecosystem is used to solve problems ranging from recommendations to news feed updates to email digesting to descriptive analytical dashboards for our members.","PeriodicalId":87344,"journal":{"name":"Proceedings. ACM-SIGMOD International Conference on Management of Data","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2013-06-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86946373","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}