This repository contains the input and output data for the research article entitled Estimating Time-Series Changes in Social Sensitivity for COVID-19 @ Twitter in Japan by Ryuichi Saito and Shinichiro Haruyama. The article was published in the Transactions of the Japanese Society for Artificial Intelligence in 2022.
The collection_data folder contains tweets collected using the Full-archive Search API of Twitter API v2 and the estimation_result folder consists of folders tweets sentiment classified by the BERT model fine-tuned by the Rakuten Dataset.