2023 Fiscal Year Final Research Report
Temporal Evolution and Segmentation of Time Series Patterns for IoT Sensor Data Analysis
| Project/Area Number |
20K11832
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Multi-year Fund |
|---|
| Section | 一般 |
|---|
| Review Section |
Basic Section 60080:Database-related
|
|---|
| Research Institution | Tokai University |
|---|
Principal Investigator |
|
|---|
| Project Period (FY) |
2020-04-01 – 2023-03-31
|
| Keywords | 時系列データマイニング / Time series motifs / Symbolic Representation / DTW Motifs / Predicative Maintenance |
|---|
| Outline of Final Research Achievements |
With the proliferation of IoT and big data, there is an increasing need for sensor data analysis, ranging from predictive maintenance in factories and public facilities to human behavior analysis and health management. However, traditional frequency analysis and time series analysis assume stationarity, making them often unsuitable for sensor data. To address this, we propose a novel time series decomposition method called Spikelet, which uses "the range of fluctuation of time series values" as a hyperparameter instead of "the time width of the time series." Additionally, we developed "grammatical motifs" to discover the structure of sensor data and "temporal evolution of motifs" to detect changes such as degradation or failure. This work has been accepted at top international conferences, including ACM-KDD and IEEE-ICDM.
|
| Free Research Field |
情報学 マルチメディア・データベース ビッグデータ分析・活用
|
| Academic Significance and Societal Importance of the Research Achievements |
四次産業革命やSociety 5.0を実現するために， IoT(Internet of Things)普及に伴って飛躍的に増加するセンサーデータを分析する基盤技術を確立する
|
