<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE article PUBLIC "-//NLM//DTD JATS (Z39.96) Journal Publishing DTD v1.3 20210610//EN" "JATS-journalpublishing1-3.dtd">
<article article-type="research-article" dtd-version="1.3" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xml:lang="ru"><front><journal-meta><journal-id journal-id-type="publisher-id">intechngu</journal-id><journal-title-group><journal-title xml:lang="ru">Вестник НГУ. Серия: Информационные технологии</journal-title><trans-title-group xml:lang="en"><trans-title>Vestnik NSU. Series: Information Technologies</trans-title></trans-title-group></journal-title-group><issn pub-type="ppub">1818-7900</issn><issn pub-type="epub">2410-0420</issn><publisher><publisher-name>НГУ</publisher-name></publisher></journal-meta><article-meta><article-id pub-id-type="doi">10.25205/1818-7900-2025-23-3-67-78</article-id><article-id custom-type="elpub" pub-id-type="custom">intechngu-326</article-id><article-categories><subj-group subj-group-type="heading"><subject>Research Article</subject></subj-group><subj-group subj-group-type="section-heading" xml:lang="ru"><subject>Статьи</subject></subj-group></article-categories><title-group><article-title>Анализ и тестирование индекса RadixSpline</article-title><trans-title-group xml:lang="en"><trans-title>Analysis and testing of the RadixSpline Learned index</trans-title></trans-title-group></title-group><contrib-group><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Стецкий</surname><given-names>И. М.</given-names></name><name name-style="western" xml:lang="en"><surname>Stetskii</surname><given-names>I. M.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Стецкий Илья Максимович, магистрант</p><p>Новосибирск</p><p> </p></bio><bio xml:lang="en"><p>Ilya M. Stetsky, Master’s Student</p><p>Novosibirsk</p></bio><email xlink:type="simple">i.stetskii@g.nsu.ru</email><xref ref-type="aff" rid="aff-1"/></contrib><contrib contrib-type="author" corresp="yes"><name-alternatives><name name-style="eastern" xml:lang="ru"><surname>Пищик</surname><given-names>Б. Н.</given-names></name><name name-style="western" xml:lang="en"><surname>Pischik</surname><given-names>B. N.</given-names></name></name-alternatives><bio xml:lang="ru"><p>Пищик Борис Николаевич, доцент</p><p>Новосибирск</p><p> </p></bio><bio xml:lang="en"><p>Boris N. Pishik, Associate Professor</p><p>Novosibirsk</p></bio><email xlink:type="simple">b.pishchik@g.nsu.ru</email><xref ref-type="aff" rid="aff-1"/></contrib></contrib-group><aff-alternatives id="aff-1"><aff xml:lang="ru"><institution>Новосибирский государственный университет</institution><country>Россия</country></aff><aff xml:lang="en"><institution>Novosibirsk State University</institution><country>Russian Federation</country></aff></aff-alternatives><pub-date pub-type="collection"><year>2025</year></pub-date><pub-date pub-type="epub"><day>10</day><month>11</month><year>2025</year></pub-date><volume>23</volume><issue>3</issue><fpage>67</fpage><lpage>78</lpage><permissions><copyright-statement>Copyright &amp;#x00A9; Стецкий И.М., Пищик Б.Н., 2025</copyright-statement><copyright-year>2025</copyright-year><copyright-holder xml:lang="ru">Стецкий И.М., Пищик Б.Н.</copyright-holder><copyright-holder xml:lang="en">Stetskii I.M., Pischik B.N.</copyright-holder><license xml:lang="ru" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>Данная работа распространяется под лицензией Creative Commons Attribution 4.0.</license-p></license><license xml:lang="en" license-type="creative-commons-attribution" xlink:href="https://creativecommons.org/licenses/by/4.0/" xlink:type="simple"><license-p>This work is licensed under a Creative Commons Attribution 4.0 License.</license-p></license></permissions><self-uri xlink:href="https://intechngu.elpub.ru/jour/article/view/326">https://intechngu.elpub.ru/jour/article/view/326</self-uri><abstract><p>Статья посвящена тестовой реализации индекса RadixSpline и исследованию его производительности при различных параметрах настойки. Индекс RadixSpline относится к категории обученного индекса и предназначен для эффективного поиска в отсортированных данных в структуре ключ – значение. Индекс строится за один проход по данным и состоит из сплайн-аппроксимации с ограниченной погрешностью (GreedySpline) и разреженного индекса (RadixTable). В работе перечислены известные обученные индексы и сравнение их с исследуемым RadixSpline, который выгодно отличается от других тем, что имеет простую структуру и малое количество параметров настройки. Проведен анализ работы RadixSpline и его частей, а также влияние параметров на производительность и объем затрачиваемой памяти. Тестирование реализации показало, что RadixSpline способен значительно сократить время поиска по сравнению с бинарным поиском. Кроме того, при других настройках можно сократить используемую память. В работе предлагаются уточнения в алгоритме RadixSpline и исследование возможности его интеграции в промышленные продукты для работы с большими данными.</p></abstract><trans-abstract xml:lang="en"><p>Recent advances in machine learning have introduced a novel approach to database indexing known as learned indexes. Among these, the RadixSpline index has emerged as a particularly promising solution for efficient search in sorted key-value data. This paper presents a comprehensive analysis and performance evaluation of the RadixSpline index, which combines spline approximation with bounded error (GreedySpline) and a sparse radix table structure. We investigate the index’s behavior under different parameter configurations, focusing on two key parameters: the spline approximation error (err) and the radix prefix length (r). Our experimental study, conducted on a 25 MB subset of the MovieLens32M dataset, demonstrates that RadixSpline can achieve up to 30% faster search times compared to traditional binary search while reducing memory usage by over 80 % compared to RocksDB’s sparse index implementation. The study makes several important contributions: (1) we provide a detailed analysis of parameter sensitivity, identifying optimal configurations for various use cases; (2) we address implementation challenges not covered in the original work, including handling of non-contiguous prefixes and missing keys; (3) we demonstrate the index’s suitability for real-time applications through its single-pass construction property. Our results confirm that RadixSpline offers significant advantages over conventional indexing methods, particularly in scenarios requiring fast searches with limited memory resources. The findings suggest strong potential for integrating RadixSpline into production database systems, with RocksDB being a particularly promising candidate for such integration.</p></trans-abstract><kwd-group xml:lang="ru"><kwd>RadixSpline</kwd><kwd>Learned Index</kwd><kwd>сплайн-аппроксимация</kwd><kwd>разреженный индекс</kwd><kwd>оптимизация поиска</kwd><kwd>производительность СУБД</kwd><kwd>машинное обучение в индексации</kwd></kwd-group><kwd-group xml:lang="en"><kwd>RadixSpline</kwd><kwd>Learned Index</kwd><kwd>spline approximation</kwd><kwd>sparse index</kwd><kwd>search optimization</kwd><kwd>DBMS performance</kwd><kwd>machine learning in indexing</kwd></kwd-group><funding-group><funding-statement xml:lang="ru">Исследование выполнено при поддержке благотворительного фонда содействия образованию и становления молодых специалистов «Высшая лига».</funding-statement><funding-statement xml:lang="en">The work was supported by the charitable foundation for the promotion of education and formation of young specialists «Высшая лига».</funding-statement></funding-group></article-meta></front><back><ref-list><title>References</title><ref id="cit1"><label>1</label><citation-alternatives><mixed-citation xml:lang="ru">Kraska T., Beutel A., Chi E.H., Dean J., Polyzotis N. The Case for Learned Index Structures // arXiv:1712.01208 [cs.DB]. 2017. DOI: 10.48550/arXiv.1712.01208</mixed-citation><mixed-citation xml:lang="en">Kraska T., Beutel A., Chi E.H., Dean J., Polyzotis N. The Case for Learned Index Structures // arXiv:1712.01208 [cs.DB]. 2017. DOI: 10.48550/arXiv.1712.01208</mixed-citation></citation-alternatives></ref><ref id="cit2"><label>2</label><citation-alternatives><mixed-citation xml:lang="ru">Kipf A., Marcus R., van Renen A., Stoian M., Kemper A., Kraska T., Neumann T. RadixSpline: A Single-Pass Learned Index // Proceedings of the Third International Workshop on Exploiting Artifi cial Intelligence Techniques for Data Management (aiDM ‘20). 2020. Art. No. 5. P. 1–5. DOI: 10.1145/3401071.3401659.</mixed-citation><mixed-citation xml:lang="en">Kipf A., Marcus R., van Renen A., Stoian M., Kemper A., Kraska T., Neumann T. RadixSpline: A Single-Pass Learned Index // Proceedings of the Third International Workshop on Exploiting Artifi cial Intelligence Techniques for Data Management (aiDM ‘20). 2020. Art. No. 5. P. 1–5. DOI: 10.1145/3401071.3401659.</mixed-citation></citation-alternatives></ref><ref id="cit3"><label>3</label><citation-alternatives><mixed-citation xml:lang="ru">Neumann T., Michel S. Smooth Interpolating Histograms with Error Guarantees // Proceedings of the 11th International Conference on Extending Database Technology (EDBT ‘08). Cardiff , UK, 2008. P. 42–53. DOI: 10.1007/978-3-540-70504-8-12.</mixed-citation><mixed-citation xml:lang="en">Neumann T., Michel S. Smooth Interpolating Histograms with Error Guarantees // Proceedings of the 11th International Conference on Extending Database Technology (EDBT ‘08). Cardiff , UK, 2008. P. 42–53. DOI: 10.1007/978-3-540-70504-8-12.</mixed-citation></citation-alternatives></ref><ref id="cit4"><label>4</label><citation-alternatives><mixed-citation xml:lang="ru">MovieLens32M Dataset [Электронный ресурс] / URL: GroupLens Research (дата обращения: 10.12.2024).</mixed-citation><mixed-citation xml:lang="en">MovieLens32M Dataset [Электронный ресурс] / URL: GroupLens Research (дата обращения: 10.12.2024).</mixed-citation></citation-alternatives></ref><ref id="cit5"><label>5</label><citation-alternatives><mixed-citation xml:lang="ru">Index Block Format [Электронный ресурс] // RocksDB Wiki. – URL: https://github.com/facebook/rocksdb/wiki/Index-Block-Format (дата обращения: 10.08.2024).</mixed-citation><mixed-citation xml:lang="en">Index Block Format [Электронный ресурс] // RocksDB Wiki. – URL: https://github.com/facebook/rocksdb/wiki/Index-Block-Format (дата обращения: 10.08.2024).</mixed-citation></citation-alternatives></ref><ref id="cit6"><label>6</label><citation-alternatives><mixed-citation xml:lang="ru">RocksDB: A Persistent Key-Value Store [Электронный ресурс] / Meta Open Source. URL: https://github.com/facebook/rocksdb (дата обращения: 10.08.2024).</mixed-citation><mixed-citation xml:lang="en">RocksDB: A Persistent Key-Value Store [Электронный ресурс] / Meta Open Source. URL: https://github.com/facebook/rocksdb (дата обращения: 10.08.2024).</mixed-citation></citation-alternatives></ref><ref id="cit7"><label>7</label><citation-alternatives><mixed-citation xml:lang="ru">O’Neil P. The Log-Structured Merge-Tree (LSM-Tree) // Acta Informatica. 1996. Vol. 33. No. 4. P. 351–385. DOI: 10.1007/s002360050048.</mixed-citation><mixed-citation xml:lang="en">O’Neil P. The Log-Structured Merge-Tree (LSM-Tree) // Acta Informatica. 1996. Vol. 33. No. 4. P. 351–385. DOI: 10.1007/s002360050048.</mixed-citation></citation-alternatives></ref><ref id="cit8"><label>8</label><citation-alternatives><mixed-citation xml:lang="ru">Marcus R., Kipf A., van Renen A., Stoian M., Misra S., Kemper A., Neumann T., Kraska T. Benchmarking Learned Indexes // Proceedings of the VLDB Endowment (PVLDB). 2021. Vol. 14. No. 1. P. 1–13. DOI: 10.14778/3421424.3421425.</mixed-citation><mixed-citation xml:lang="en">Marcus R., Kipf A., van Renen A., Stoian M., Misra S., Kemper A., Neumann T., Kraska T. Benchmarking Learned Indexes // Proceedings of the VLDB Endowment (PVLDB). 2021. Vol. 14. No. 1. P. 1–13. DOI: 10.14778/3421424.3421425.</mixed-citation></citation-alternatives></ref><ref id="cit9"><label>9</label><citation-alternatives><mixed-citation xml:lang="ru">Li Z., Chan T. N., Yiu M. L., Jensen C. S. A Survey of Learned Indexes for the Multi-dimensional Space // arXiv:2403.06456 [cs.DB]. 2024. DOI: 10.48550/arXiv.2403.06456</mixed-citation><mixed-citation xml:lang="en">Li Z., Chan T. N., Yiu M. L., Jensen C. S. A Survey of Learned Indexes for the Multi-dimensional Space // arXiv:2403.06456 [cs.DB]. 2024. DOI: 10.48550/arXiv.2403.06456</mixed-citation></citation-alternatives></ref><ref id="cit10"><label>10</label><citation-alternatives><mixed-citation xml:lang="ru">Ferragina P., Vinciguerra G. A Critical Analysis of Recursive Model Indexes // arXiv:2106.16166 [cs.DB]. 2021. DOI: 10.48550/arXiv.2106.16166</mixed-citation><mixed-citation xml:lang="en">Ferragina P., Vinciguerra G. A Critical Analysis of Recursive Model Indexes // arXiv:2106.16166 [cs.DB]. 2021. DOI: 10.48550/arXiv.2106.16166</mixed-citation></citation-alternatives></ref><ref id="cit11"><label>11</label><citation-alternatives><mixed-citation xml:lang="ru">Ding J., Minhas U. F., Yu J., Wang C., Li Y., Zhang H., Chandramouli B., Gehrke J., Kossmann D., Lomet D., Kraska T. Towards Practical Learned Indexing (Extended Abstracts) // arXiv:2108.05117 [cs.DB]. 2021. DOI: 10.48550/arXiv.2108.05117</mixed-citation><mixed-citation xml:lang="en">Ding J., Minhas U. F., Yu J., Wang C., Li Y., Zhang H., Chandramouli B., Gehrke J., Kossmann D., Lomet D., Kraska T. Towards Practical Learned Indexing (Extended Abstracts) // arXiv:2108.05117 [cs.DB]. 2021. DOI: 10.48550/arXiv.2108.05117</mixed-citation></citation-alternatives></ref><ref id="cit12"><label>12</label><citation-alternatives><mixed-citation xml:lang="ru">Wu J., Zhang Y., Chen S., Wang J., Chen Y., Xing C. LSI: A Learned Secondary Index Structure // arXiv:2205.05769 [cs.DB]. 2022. DOI: 10.48550/arXiv.2205.05769</mixed-citation><mixed-citation xml:lang="en">Wu J., Zhang Y., Chen S., Wang J., Chen Y., Xing C. LSI: A Learned Secondary Index Structure // arXiv:2205.05769 [cs.DB]. 2022. DOI: 10.48550/arXiv.2205.05769</mixed-citation></citation-alternatives></ref><ref id="cit13"><label>13</label><citation-alternatives><mixed-citation xml:lang="ru">Hadian A., Heinis T. Shift-Table: A Low-Latency Learned Index for Range Queries using Model Correction // arXiv:2103.13160 [cs.DB]. 2021. DOI: 10.48550/arXiv.2103.13160</mixed-citation><mixed-citation xml:lang="en">Hadian A., Heinis T. Shift-Table: A Low-Latency Learned Index for Range Queries using Model Correction // arXiv:2103.13160 [cs.DB]. 2021. DOI: 10.48550/arXiv.2103.13160</mixed-citation></citation-alternatives></ref><ref id="cit14"><label>14</label><citation-alternatives><mixed-citation xml:lang="ru">Vaidya K., Kraska T. Partitioned Learned Bloom Filter // Proceedings of the VLDB Endowment. 2020. Vol. 13. No. 12. P. 3298–3311. DOI: 10.14778/3415478.3415545</mixed-citation><mixed-citation xml:lang="en">Vaidya K., Kraska T. Partitioned Learned Bloom Filter // Proceedings of the VLDB Endowment. 2020. Vol. 13. No. 12. P. 3298–3311. DOI: 10.14778/3415478.3415545</mixed-citation></citation-alternatives></ref><ref id="cit15"><label>15</label><citation-alternatives><mixed-citation xml:lang="ru">Spector B., Marcus R., van Renen A., Stoian M., Kemper A., Kraska T., Neumann T. Bounding the Last Mile: Effi cient Learned String Indexing // arXiv:2111.14905 [cs.DB]. 2021. DOI: 10.48550/arXiv.2111.14905</mixed-citation><mixed-citation xml:lang="en">Spector B., Marcus R., van Renen A., Stoian M., Kemper A., Kraska T., Neumann T. Bounding the Last Mile: Effi cient Learned String Indexing // arXiv:2111.14905 [cs.DB]. 2021. DOI: 10.48550/arXiv.2111.14905</mixed-citation></citation-alternatives></ref><ref id="cit16"><label>16</label><citation-alternatives><mixed-citation xml:lang="ru">Crotty A. Hist-Tree: Those Who Ignore It Are Doomed to Learn [Электронный ресурс] // CIDR 2021. URL: https://cs.brown.edu/people/acrotty/pubs/cidr2021-paper20.pdf (дата обращения: 02.02.2025).</mixed-citation><mixed-citation xml:lang="en">Crotty A. Hist-Tree: Those Who Ignore It Are Doomed to Learn [Электронный ресурс] // CIDR 2021. URL: https://cs.brown.edu/people/acrotty/pubs/cidr2021-paper20.pdf (дата обращения: 02.02.2025).</mixed-citation></citation-alternatives></ref></ref-list><fn-group><fn fn-type="conflict"><p>The authors declare that there are no conflicts of interest present.</p></fn></fn-group></back></article>
