Science that makes the news does.

These are general notes summarizing comments from audience members and speakers for this session:

• Taking pdfs extracting text and semantically marking them up, hyperlinking reference lists to their source articles.
• Mendeley http://www.mendeley.com/
• http://www.mendeley.com/research-papers/computer-and-information-science/
• Citations have no meaning the way we use them now.
• Most papers aren’t cited, long tail graph (zitiny  ziphian curve graph?)
• Citation Graph, collaborative filtering,.
• Citations need context. How will that happen? Need to know why something was cited (ie. Disagreement).
• We don’t disagree with papers, we disagree with claims made in papers. Why can’t a citation point at a place in the article?
• Question: Is reference extraction beyond the scope of any non-commercial company? Mendeley is a company.
•                Answer: There’s no single source of open bibliographic data.
• Sage, SocialSciences,Crossref, webofscience, google scholar, etc – how useful are these sites. They provide metrics and recommendations, but not much more than going directly to the journal.
• Social Networking in article recommendations, connect content to people, conversations around papers, systems don’t encourage conversations, people don’t want to participate.
• Criticism of Mendeley: algorithms shoudl be open, academics should be able to define their own algorithms
• Mendeley's plan is to extract reference data and make it publicly available and machine readable.
• Criticism: academics need an open bibliographic data set.
• Need to explain type of citation: positive vs. negative citations, valence terms, sentiment analysis/machine vs. human curated
• Ontologies don't capture all reasons someone is citing something (ie. "Cited because I work for this journal." "Cited because Darwin will make you think I'm smart." "Cited because teacher required five citations.")
• References are separated into their own section, removing them from the text. Unlike links, which are immediate.
• Peer Review: example of a reviewer rejecting a paper because it didn't cite his own paper.
• People need to make use of the REL attribute in HREF tags.
• Citations can be used in a tribal sens, citing people in our camp and excluding others.
• Description of citations as "frozen footprints in the snow"
• Why do we need 1,000 citation styles?
• Librarian: Questions about citations styles from students are constant and frustrating.
• Orchid: cross-company effort to standardize citations.
• Let people write citations however they want, but add an identifying number.
• Mendeley is developing an open-source citation style editor.

The page rank algorithm is very similar to an algorithm for ranking research papers by counting the number of other papers referencing them.