5+ Years of Multiple Time Series Data of Hourly PM2.5 Recordings Gathered from Various Sensors Located throughout Japan (1-1-2018 to 25-4-2023)

Air pollution is the major cause for many respiratory related deaths in Japan. The Ministry of Environment Japan has setup a nation-wide sensor network to monitor air pollution throughout Japan. This network is called Atmospheric Environmental Regional Observation System (AEROS). More information on this is available at https://soramame.env.go.jp/

Here, we are providing the hourly recordings of PM2.5 values by 1000+ sensors from 1-1-2018 to 25-4-2023 (5+ years). Please consider the value 9999 means missing data. According to Japanese norms, values greater than 15 ppm3 is consider unhealthy.

We are providing three files:
a) station_info.csv : contains the station_id, address, and lat/long of the station
b) ETL_DATA_final.csv: Contains the PM2.5 hourly data of various sensors.
format of this data is: rowNumber,timestamp,sensor_1 values, sensor_2 values, ...
c) Soramame-key.pdf: Explains the process to convert this columnar multiseries data frame into a transactional database and extract patterns.

Citation:
Rage Uday Kiran, Pamalla Veena, Penugonda Ravikumar, Bathala Venus Vikranth Raj, Minh-Son Dao, Koji Zettsu, Sai Chithra Bommisetti:
HDSHUI-miner: a novel algorithm for discovering spatial high-utility itemsets in high-dimensional spatiotemporal databases. Appl. Intell. 53(8): 8536-8561 (2023)

大气污染是导致日本诸多呼吸道相关死亡病例的首要诱因。日本环境省已搭建覆盖全国的大气污染监测传感器网络，该网络被命名为大气环境区域观测系统（Atmospheric Environmental Regional Observation System, AEROS）。更多相关详情可访问https://soramame.env.go.jp/。

本次我们公开了2018年1月1日至2023年4月25日（共计5年以上）期间，1000余台传感器的每小时PM2.5浓度监测数据。请注意，数据集中取值为9999的项代表缺失值。根据日本相关标准，浓度高于15 ppm³的PM2.5数据被视为不健康水平。

本次公开三类数据文件：
a) station_info.csv：包含各监测站点的站点ID、地址及经纬度坐标；
b) ETL_DATA_final.csv：收录各传感器的PM2.5每小时监测数据，其数据格式为：行号、时间戳、传感器1数值、传感器2数值……；
c) Soramame-key.pdf：详细阐述了如何将该列式多序列数据帧转换为事务型数据库并提取模式的具体流程。

引用文献：
Rage Uday Kiran, Pamalla Veena, Penugonda Ravikumar, Bathala Venus Vikranth Raj, Minh-Son Dao, Koji Zettsu, Sai Chithra Bommisetti:
HDSHUI-miner：一种面向高维时空数据库的空间高效用项集挖掘新型算法。Appl. Intell. 53(8): 8536-8561 (2023)