The dataset contains metadata for more than 50,000 tweets on Christmas Eve and Christmas Day 2016. The first field is the ID of the tweet and the second field is the HTML metadata. The researchers obtained this data using a web crawler written in Java. It is recommended to use BeautifulSoup or other libraries to parse this data and extract information from each tweet. The dataset could help observe some kind of demographic data about the usage of emojis in tweets, the type of emotions associated with Christmas, or the age or nationality of active Twitter users during Christmas.