The Belt and Road Initiative on Twitter: An Annotated Dataset

This repository stores a collection of 500,662 posts (tweets) and 714,701 reposting threads (retweets) related to the BRI on Twitter. The dataset was collected through the Twitter API using the terms: “belt and road”, “one belt one road”, ‘new silk road”, “maritime silk road”, and “silk road economic belt”, which included the phrases and their hashtag format to download data. The time series of the dataset is from 7 September 2013 to 30 November 2021. Furthermore, the dataset has been annotated in terms of languages, sentiment, and geopolitical entities by using computational analytics in language detection, neural machine translation, lexicon-based sentiment analysis, and web automation. In order to facilitate insights discovery, we classified the dataset into four databases in which can be analyzed separately and used by researchers related to various fields, such as social science, network science, sociology, and others to study the dynamic activities on Twitter, trends on keywords, users’ characteristics, as well as the social networks on citing patterns.