| The increasing availability of big scholarly data and the widespread use of social media have created opportunities for the study of scholarly communication. This dissertation concerns about scholarly communication across three spheres, namely the scientific community, the patented technologies domain, and online social media. The first work examines how scholarly papers are cited over time within the scientific community by focusing on the delayed recognition phenomenon in science. The analysis points out that such phenomenon is not exceptional, as previous literature has repeatedly suggested. I then study factors linked to citation counts of papers, showing their limited ability to explain the variation in citations. An evolution model is also proposed to reproduce the tie strength distribution in scientific collaboration networks. As the domains where paper can be cited are not restricted to the scientific community, next, I look at citations of papers from the patented technologies domain. I present a comparative study on how papers are cited by U.S. patents and by other papers over time, using a newly-created dataset that links non-patent references made by U.S. patents to biomedical papers. My analysis shows that the two types of citations exhibit distinct temporal variations. Finally, while metrics derived from social medial have been argued to provide broader impact and advocated for research evaluation, actors who participate in conversations about science on social media are largely unknown. Motivated by this issue, an approach is proposed to identify scientists on Twitter. The analysis shows that Twitter has been employed by scholars across the disciplinary spectrum and scholarly communication on social media rarely happens across disciplinary boundaries. |
