研究課題/領域番号 |
22H03573
|
研究種目 |
基盤研究(B)
|
配分区分 | 補助金 |
応募区分 | 一般 |
審査区分 |
小区分60060:情報ネットワーク関連
|
研究機関 | 東京大学 |
研究代表者 |
FAN ZIPEI 東京大学, 空間情報科学研究センター, 客員研究員 (70835397)
|
研究期間 (年度) |
2022-04-01 – 2025-03-31
|
研究課題ステータス |
交付 (2023年度)
|
配分額 *注記 |
8,190千円 (直接経費: 6,300千円、間接経費: 1,890千円)
2023年度: 2,600千円 (直接経費: 2,000千円、間接経費: 600千円)
2022年度: 2,990千円 (直接経費: 2,300千円、間接経費: 690千円)
|
キーワード | Crowdsensing / Privacy preserving / Trustworthy AI / Graph Neural Network / Privacy Preserving / Data Valuation / Federated Learning / Spatiotemporal Modeling / Data Privacy |
研究開始時の研究の概要 |
Mobile Crowdsensing is an emerging participatory sensing paradigm that collects participants’ surrounding information via mobile devices. In this research, we aim at addressing the data privacy, data reliability and data valuation issues in the mobile crowdsensing. Data privacy of the participants will be protected via a federated learning framework, data reliability problem will be addressed via a spatiotemporal filtering approach on heterogeneous graph neural network, and the data from each participant will be valuated via a Shapley value approach on the basis of cooperative game theory.
|
研究実績の概要 |
This year, I initiated the research on heterogeneous graph neural network based privacy preserving crowdsensing system. In this year, I have achieved several publications: one paper on using heterogeneous graph neural network for location-based services and published in CIKM 2022 and WWWJ 2022, and two papers on federated learning based crowdsensing system are published in IOTJ 2022, heterogeneous graph neural network based methods for crowdsensing way of indoor localization is published in IOTJ 2023, and shapley-based data valuation method for traffic prediction in Sensors 2022.
|
現在までの達成度 (区分) |
現在までの達成度 (区分)
1: 当初の計画以上に進展している
理由
This year, I have moved smoothly on the developing the heterogeneous graph neural network privacy preserving crowdsensing system. Although the complete system is not finished yet, but most of the core components have been researched and prototypes have been developed.
|
今後の研究の推進方策 |
In the second year of this project, I planned to further conduct relevant research on these technics as planned, and also try to have a better integration of these technics to model a crowdsensing system that is both practical, privacy preserving and fair.
|