Knowledge discovery from compressed stream data

2017 Fiscal Year Final Research Report

• Project/Area Number: 26280088
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Intelligent informatics
• Research Institution: Kyushu Institute of Technology
• Principal Investigator: Sakamoto Hiroshi 九州工業大学, 大学院情報工学研究院, 教授 (50315123)
• Co-Investigator(Kenkyū-buntansha): 山際 伸一 筑波大学, システム情報工学研究科(系), 准教授 (10574725) ; 榎田 修一 九州工業大学, 大学院情報工学研究院, 教授 (40346862)
• Co-Investigator(Renkei-kenkyūsha): TABEI Yasuo 理化学研究所, 革新知能統合研究センター, ユニットリーダー (20589824)
• Project Period (FY): 2014-04-01 – 2017-03-31
• Keywords: ストリームデータ / 圧縮伝送 / FPGA / 可逆圧縮

Outline of Final Research Achievements

The aim of this research is to produce a novel technology for stream data compression realizing the real time processing of stream data to obtain deep knowledge hidden big data stream. In the previous studies, because of the tradeoff between compression ratio and memory consumption, either algorithm requires so huge memory to load whole input or input data must be decomposed into a sequence of small segments so that the algorithm can load it into the restricted memory. In this proposal, we can minimize the nonnegligible tradeoff, and we can expand the throughput of the network constructed by the compression-transmission-decompression algorithm implemented by FPGA. We propose and demonstrate these outcomes to the real world via patent applications, research articles, and industrial demonstrations.

Free Research Field

データ圧縮