研究課題/領域番号 |
21K12034
|
研究種目 |
基盤研究(C)
|
配分区分 | 基金 |
応募区分 | 一般 |
審査区分 |
小区分61030:知能情報学関連
|
研究機関 | 会津大学 |
研究代表者 |
Rage Uday・Kiran 会津大学, コンピュータ理工学部, 准教授 (20874324)
|
研究分担者 |
是津 耕司 国立研究開発法人情報通信研究機構, ユニバーサルコミュニケーション研究所統合ビッグデータ研究センター, 研究センター長 (40415857)
|
研究期間 (年度) |
2021-04-01 – 2025-03-31
|
研究課題ステータス |
交付 (2022年度)
|
配分額 *注記 |
4,030千円 (直接経費: 3,100千円、間接経費: 930千円)
2024年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2023年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2022年度: 780千円 (直接経費: 600千円、間接経費: 180千円)
2021年度: 1,690千円 (直接経費: 1,300千円、間接経費: 390千円)
|
キーワード | Big data / high utility patterns / spatial information / data mining / Pattern mining / Air pollution analytics |
研究開始時の研究の概要 |
“Mining time series data” is one of the top-10 challenges in data mining. This research aims to tackle this challenging problem of great importance by proposing a mathematical model to uncover periodic spatial patterns in irregular spatiotemporal big data. We will deliver a mathematical model and software programs to uncover interesting patterns in spatiotemporal big data. Our deliverables will be “open-sourced” to foster R&D on data mining.
|
研究実績の概要 |
High utility itemset mining is an important knowledge discovery in data mining. Exicisting studies ignored the spatial information of the items in the database and tried to find hidden patterns. We have observed that ignoring the spatial information results either in missing useful information or generating suprious information that is not useful to the experts. In this year, we have tested our cliam, which is important to capture the spatial information of the items to find patterns that have high value. The work was published in Applied Intelligence Journal.
|
現在までの達成度 (区分) |
現在までの達成度 (区分)
2: おおむね順調に進展している
理由
We were successful in downloading the air pollution data of various sensors in Japan, USA, and other countries. We were able to test the distribution of characteritics of the data. We found that existing high utility itemset mining algorithms were inadequate to find spatially interesting patterns as they ignored the spatial information. We proposed a new model and an efficient algorithm to discover spatially interesting patterns having high value. Our work also published in Applied Intelligence journal.
|
今後の研究の推進方策 |
1. This year we will model the air pollution has uncertain data generated by the sensors and study the current limitations. Next, we will analyze the results and make claim, followed by introducing a new model and algorithm to find spatially interesting patterns in uncertain data.
2. We will also study how to model air pollution as fuzzy dataset to extract useful information
|