Annotating patents with Medline MeSH codes via citation mapping.

Both patents and Medline are important document collections for discovering new relationships between chemicals and biology, searching for prior art for patent applications and retrieving background knowledge for current research activities. Finding relevance to a topic within patents is often made difficult by poor categorization, badly written descriptions, and even intentional obfuscation. Unlike patents, the Medline corpus has Medical Subject Heading (MeSH) keywords manually added to their articles, giving a medically relevant taxonomy to the 18 million article abstracts.

Who gets acknowledged: measuring scientific contributions through automatic acknowledgment indexing.

Acknowledgements in research publications, like citations, indicate influential contributions to scientific work. However, acknowledgements are different from citations; whereas citations are formal expressions of debt, acknowledgements are arguably more personal, singular, or private expressions of appreciation and contribution. Furthermore, many sources of research funding expect researchers to acknowledge any support that contributed to the published work.