Airware-Haikou

The AirWare-Haikou dataset, POI data, and public complaint corpus data utilised in the study. The structure of the repository consists of three folders: complaint corpus data, POI data, and AQMD data, which are stored in CSV or XLS format for readers to read and process. (1) MTSAM dataset: The folder contains the file station_lon.csv, which records the names, addresses, and geographic coordinates (latitude and longitude) of 95 monitoring stations. The subfolders C1-C4 store four sub-datasets formed by clustering air quality time series by station. Each sub-dataset includes three files: train.csv, test.csv, and test_label.csv, which are used for training and testing. (2) POI data: This folder contains a subfolder and a file. The subfolder stores POI points categorized by type, while the file provides all the initial POI points. (3) Complaints Corpus data: This folder contains a single file, Complaint_corpus.csv, with approximately 2,000 records. The modeling algorithm (DsRL-Net) involved in this study is available to researchers who need it upon request to the authors. The readme.md file in the code package describes how to run the code for the researchers.