benchmark set to speed up extremely large scale data analysis applications

2019 Fiscal Year Final Research Report

• Project/Area Number: 15K21423
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Software; High performance computing
• Research Institution: Meiji University
• Principal Investigator: Akioka Sayaka 明治大学, 総合数理学部, 専任教授 (90333533)
• Project Period (FY): 2015-04-01 – 2020-03-31
• Keywords: ベンチマーク

Outline of Final Research Achievements

This project aims to build a benchmark suite, which is expected to help speeding up data stream analysis, with special focus on the extremely large scale input data. More concretely, categorization of streaming analysis is expected to introduce characteristic behaviors of the programs, and representative examples for each category are expected to form a benchmark suite for this community. After the period of this research project, the benchmark suite could not be completed. Struggles during the research period brought several directions for the further research. For example, input data has much influence than expected, and the variety of the pattern is enormous. Extraction of dominative characteristics in input data is a hard problem, and speculative execution is the current solution. Modeling of input data is unavoidable for modeling of programs in this area.

Free Research Field

並列分散処理

Academic Significance and Societal Importance of the Research Achievements

大規模データ解析は現在でも重要で注目を浴びているアプリケーションであり、中でもストリーム解析はリアルタイムに大規模なデータ解析を行う上で不可欠である。このように社会的需要が高いアプリケーションを高速化することは、新たな応用分野の開拓や新しい知見を得るために不可欠であるが、抜本的な解決法は発見されていない。本研究でも問題解決に至ることはできなかったが、問題解決に向けて取り組むべき問題を以前より具体的にすることはできた。