{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,3]],"date-time":"2025-11-03T01:23:59Z","timestamp":1762133039622,"version":"build-2065373602"},"reference-count":29,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2025,1,13]],"date-time":"2025-01-13T00:00:00Z","timestamp":1736726400000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2025,1,13]],"date-time":"2025-01-13T00:00:00Z","timestamp":1736726400000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["Distrib Parallel Databases"],"published-print":{"date-parts":[[2025,12]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>The rapid growth of large-scale machine learning (ML) models has led numerous commercial companies to utilize ML models for generating predictive results to help business decision-making. As two primary components in traditional predictive pipelines, data processing, and model predictions often operate in separate execution environments, leading to redundant engineering and computations. Additionally, the diverging mathematical foundations of data processing and machine learning hinder cross-optimizations by combining these two components, thereby overlooking potential opportunities to expedite predictive pipelines. In this paper, we propose an operator fusion method based on GPU-accelerated linear algebraic evaluation of relational queries. Our method leverages linear algebra computation properties to merge operators in machine learning predictions and data processing, significantly accelerating predictive pipelines by up to 317x. We perform a complexity analysis to deliver quantitative insights into the advantages of operator fusion, considering various data and model dimensions. Furthermore, we extensively evaluate linear algebra query processing and operator fusion utilizing the widely-used Star Schema and TPC-DI benchmarks. Through comprehensive evaluations, we demonstrate the effectiveness and potential of our approach in improving the efficiency of data processing and machine learning workloads on modern hardware.<\/jats:p>","DOI":"10.1007\/s10619-024-07451-7","type":"journal-article","created":{"date-parts":[[2025,1,12]],"date-time":"2025-01-12T22:16:43Z","timestamp":1736720203000},"update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":1,"title":["Accelerating machine learning queries with linear algebra query processing"],"prefix":"10.1007","volume":"43","author":[{"given":"Wenbo","family":"Sun","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Asterios","family":"Katsifodimos","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Rihan","family":"Hai","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"297","published-online":{"date-parts":[[2025,1,13]]},"reference":[{"key":"7451_CR1","doi-asserted-by":"publisher","unstructured":"Amossen, R.R., Pagh, R.: Faster Join-Projects and Sparse Matrix Multiplications. In: ICDT 2009, pp. 121\u2013126. Association for Computing Machinery, New York (2009). https:\/\/doi.org\/10.1145\/1514894.1514909","DOI":"10.1145\/1514894.1514909"},{"key":"7451_CR2","first-page":"362","volume":"2013","author":"C Balkesen","year":"2013","unstructured":"Balkesen, C., Teubner, J., Alonso, G., et al.: Main-memory hash joins on multi-core CPUs: tuning to the underlying hardware. ICDE 2013, 362\u2013373 (2013)","journal-title":"ICDE"},{"key":"7451_CR3","unstructured":"BlazingDB: BlazingSQL. https:\/\/github.com\/BlazingDB\/blazingsql (2020)"},{"key":"7451_CR4","first-page":"578","volume":"2018","author":"T Chen","year":"2018","unstructured":"Chen, T., Moreau, T., Jiang, Z., et al.: TVM: an automated End-to-End optimizing compiler for deep learning. OSDI 2018, 578\u2013594 (2018)","journal-title":"OSDI"},{"key":"7451_CR5","doi-asserted-by":"publisher","unstructured":"Cheng, Z., Koudas, N., Zhang, Z., et\u00a0al.: Efficient construction of nonlinear models over normalized data. In: 2021 IEEE 37th International Conference on Data Engineering (ICDE), pp. 1140\u20131151 (2021). https:\/\/doi.org\/10.1109\/ICDE51399.2021.00103","DOI":"10.1109\/ICDE51399.2021.00103"},{"key":"7451_CR6","first-page":"1213","volume":"2020","author":"S Deep","year":"2020","unstructured":"Deep, S., Hu, X., Koutris, P.: Fast join project query evaluation using matrix multiplication. SIGMOD 2020, 1213\u20131223 (2020)","journal-title":"SIGMOD"},{"key":"7451_CR7","doi-asserted-by":"crossref","unstructured":"Friedman, J.H.: Greedy function approximation: a gradient boosting machine. Ann. Stat. 29(5), 1189\u20131232 (2001). http:\/\/www.jstor.org\/stable\/2699986","DOI":"10.1214\/aos\/1013203451"},{"key":"7451_CR8","unstructured":"Gandhi, A., Asada, Y., Fu, V., et\u00a0al.: The tensor data platform: towards an ai-centric database system. In: CIDR 2023 (2023)"},{"key":"7451_CR9","doi-asserted-by":"crossref","unstructured":"Ghiran, A.M., Buchmann, R.A.: The model-driven enterprise data fabric: a proposal based on conceptual modelling and knowledge graphs. In: Douligeris, C., Karagiannis, D., Apostolou, D. (eds.) Knowledge Science, pp. 572\u2013583. Springer, Engineering and Management (2019)","DOI":"10.1007\/978-3-030-29551-6_51"},{"key":"7451_CR10","doi-asserted-by":"crossref","unstructured":"Hai, R., Koutras, C., Ionescu, A., et\u00a0al.: Amalur: data integration meets machine learning. In: ICDE 2023, p To appear (2023)","DOI":"10.1109\/ICDE55515.2023.00301"},{"key":"7451_CR11","doi-asserted-by":"crossref","unstructured":"He, D., Nakandala, S.C., Banda, D., et\u00a0al.: Query Processing on Tensor Computation Runtimes. vol\u00a015, pp. 2811\u20132825. VLDB Endowment (2022)","DOI":"10.14778\/3551793.3551833"},{"key":"7451_CR12","unstructured":"Heavy.ai: HeavyDB. https:\/\/github.com\/heavyai\/heavydb (2022)"},{"key":"7451_CR13","first-page":"1360","volume":"2022","author":"YC Hu","year":"2022","unstructured":"Hu, Y.C., Li, Y., Tseng, H.W.: Tcudb: accelerating database with tensor processors. SIGMOD 2022, 1360\u20131374 (2022)","journal-title":"SIGMOD"},{"key":"7451_CR14","doi-asserted-by":"crossref","unstructured":"Huang, Z., Chen, S.: Density-optimized intersection-free mapping and matrix multiplication for join-project operations. vol\u00a015, pp. 2244\u20132256. VLDB Endowment (2022)","DOI":"10.14778\/3547305.3547326"},{"key":"7451_CR15","doi-asserted-by":"crossref","unstructured":"Hutchison, D., Howe, B., Suciu, D.: LaraDB: a minimalist kernel for linear and relational algebra computation. In: Proceedings of the 4th ACM SIGMOD Workshop on Algorithms and Systems for MapReduce and Beyond, BeyondMR\u201917 (2017)","DOI":"10.1145\/3070607.3070608"},{"key":"7451_CR16","unstructured":"Kimball, R., Ross, M.: The Data Warehouse Toolkit: The Definitive Guide to Dimensional Modeling, 3rd edn. Wiley Publishing (2013)"},{"issue":"10","key":"7451_CR17","doi-asserted-by":"publisher","first-page":"1797","DOI":"10.14778\/3467861.3467869","volume":"14","author":"D Koutsoukos","year":"2021","unstructured":"Koutsoukos, D., Nakandala, S., Karanasos, K., et al.: Tensors: an abstraction for general data processing. Proc. VLDB Endow. 14(10), 1797\u20131804 (2021)","journal-title":"Proc. VLDB Endow."},{"key":"7451_CR18","doi-asserted-by":"publisher","first-page":"263","DOI":"10.1016\/j.procs.2021.12.013","volume":"196","author":"IA Machado","year":"2022","unstructured":"Machado, I.A., Costa, C., Santos, M.Y.: Data mesh: concepts and principles of a paradigm shift in data architectures. Procedia Comput. Sci. 196, 263\u2013271 (2022)","journal-title":"Procedia Comput. Sci."},{"key":"7451_CR19","unstructured":"Nakandala, S., Saur, K., Yu, G.I., et\u00a0al.: A tensor compiler for unified machine learning prediction serving. In: OSDI 2020. USENIX Association (2020)"},{"key":"7451_CR20","unstructured":"Okuta, R., Unno, Y., Nishino, D., et\u00a0al.: CuPy: a NumPy-compatible library for NVIDIA GPU calculations. In: NIPS 2017 Workshop: LearningSys (2017)"},{"key":"7451_CR21","doi-asserted-by":"crossref","unstructured":"O\u2019Neil, P., O\u2019Neil, E., Chen, X., et\u00a0al.: The star schema benchmark and augmented fact table indexing. In: Performance Evaluation and Benchmarking: 1st TPC Technology Conference, TPCTC 2009, pp. 237\u2013252. Springer (2009)","DOI":"10.1007\/978-3-642-10424-4_17"},{"key":"7451_CR22","first-page":"587","volume":"2022","author":"K Park","year":"2022","unstructured":"Park, K., Saur, K., Banda, D., et al.: End-to-end optimization of machine learning prediction queries. SIGMOD 2022, 587\u2013601 (2022)","journal-title":"SIGMOD"},{"key":"7451_CR23","unstructured":"Paszke, A., Gross, S., Massa, F., et\u00a0al.: Pytorch: an imperative style, high-performance deep learning library. In: NeurIPS 2019 (2019)"},{"issue":"13","key":"7451_CR24","doi-asserted-by":"publisher","first-page":"1367","DOI":"10.14778\/2733004.2733009","volume":"7","author":"M Poess","year":"2014","unstructured":"Poess, M., Rabl, T., Jacobsen, H.A., et al.: Tpc-di: the first industry benchmark for data integration. Proc. VLDB Endow. 7(13), 1367\u20131378 (2014). https:\/\/doi.org\/10.14778\/2733004.2733009","journal-title":"Proc. VLDB Endow."},{"key":"7451_CR25","doi-asserted-by":"publisher","unstructured":"Psallidas, F., Zhu, Y., Karlas, B., et\u00a0al.: Data science through the looking glass: analysis of millions of GitHub notebooks and ML.NET Pipelines. SIGMOD Rec 51(2), 30\u201337. https:\/\/doi.org\/10.1145\/3552490.3552496 (2022)","DOI":"10.1145\/3552490.3552496"},{"key":"7451_CR26","unstructured":"Rapidsai: cuDF. https:\/\/github.com\/rapidsai\/cudf (2022)"},{"key":"7451_CR27","doi-asserted-by":"crossref","unstructured":"Sun, W., Katsifodimos, A., Hai, R.: An empirical performance comparison between matrix multiplication join and hash join on GPUs. In: ICDE 2023 Workshop: HardBD & Activep (to appear) (2023)","DOI":"10.1109\/ICDEW58674.2023.00034"},{"key":"7451_CR28","unstructured":"Transaction Processing Performance Council TPC Benchmark H. http:\/\/tpc.org\/tpc_documents_current_versions\/pdf\/tpc-h_v2.18.0.pdf (2018)"},{"issue":"1","key":"7451_CR29","doi-asserted-by":"publisher","first-page":"2","DOI":"10.1145\/1077464.1077466","volume":"1","author":"R Yuster","year":"2005","unstructured":"Yuster, R., Zwick, U.: Fast sparse matrix multiplication. ACM Trans. Algorithms 1(1), 2\u201313 (2005). https:\/\/doi.org\/10.1145\/1077464.1077466","journal-title":"ACM Trans. Algorithms"}],"container-title":["Distributed and Parallel Databases"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10619-024-07451-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s10619-024-07451-7\/fulltext.html","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s10619-024-07451-7.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,11,3]],"date-time":"2025-11-03T01:19:19Z","timestamp":1762132759000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s10619-024-07451-7"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1,13]]},"references-count":29,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2025,12]]}},"alternative-id":["7451"],"URL":"https:\/\/doi.org\/10.1007\/s10619-024-07451-7","relation":{},"ISSN":["0926-8782","1573-7578"],"issn-type":[{"type":"print","value":"0926-8782"},{"type":"electronic","value":"1573-7578"}],"subject":[],"published":{"date-parts":[[2025,1,13]]},"assertion":[{"value":"3 December 2024","order":1,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"13 January 2025","order":2,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}},{"order":1,"name":"Ethics","group":{"name":"EthicsHeading","label":"Declarations"}},{"value":"The authors declare no Conflict of interest.","order":2,"name":"Ethics","group":{"name":"EthicsHeading","label":"Conflict of interest"}}],"article-number":"8"}}