Dataset of tweets, used to detect hazardous events at the Baths of Diocletian site in Rome

This dataset is composed by 276865 tweets, extracted from the Twitter stream, in the period from May 2018 to May 2019. Each element is characterized by the ID, that permits to retrieve the tweet from the stream, the text content, the GPS information (if it is provided or not), the localization, and the time. The last attribute of the elements of the dataset is the label associated to the tweet after the process of event detection. If the tweet has been identified as containing useful information of an hazardous event at the Baths of Diocletian site in Rome, the last attribute (Generated  Useful Info at BOD) is "yes", otherwise its value is "no".