摘要：A recent survey from Pew Research Center (NW, Washington & Inquiries 2018) found that over 44 million people receive science-related information from social media channels to which they subscribe. These include a variety of topics such as new discoveries in health sciences as well as “news you can use” information with practical tips (p. 3). Social and news media attention to scientific publications has been tracked for almost a decade by several platforms which aggregate all public mentions of and interactions with scientific publications. Since the amount of comments, shares, and discussions of scientific publications can reach an audience of thousands, understanding the overall “sentiment” towards the published research is a mammoth task and typically involves merely reading as many posts, shares, and comments as possible. This paper describes an initiative to track and provide sentiment analysis to large social and news media mentions and label them as “positive”, “negative” or “neutral”. Such labels will enable an overall understanding of the content that lies within these social and news media mentions of scientific publications.
