WComplexity application Web based 'leaderboard' of simple academic writing competition.
Created by: russelljjarvis
Hi WCTeam
I realized that one cool application of the WC project could be to try to foster competition between scientific writers. It's possible to use WC metrics to evaluate present and scientific writings. Some other fields are advocating for such competitions in order to promote best practices.
To this end, I mined my PI and co-PI on Google Scholar citations and compared them to two benchmark references: 'The readability of scientific texts is decreasing over time', and the 'http://splasho.com/upgoer5/library.php' xkcd upgoer5 editor library corpus (it only permits 10,000 of the most common English words). For good measure, I also added in one of Peter Marting's recent publication abstracts.
https://scholar.google.com/citations?user=GzG5kRAAAAAJ&hl=en&oi=sra https://scholar.google.com/citations?user=xnsDhO4AAAAJ&hl=en&oe=ASCII&oi=sra
I then scored all 5 authors, based on a metric, that takes into account: reading grade level, concision/redundancy, and objectivity.
The competition results confirmed a priori beliefs about text quality: 'The readability of scientific texts is decreasing over time' came first, the upgoer5 library came second, and then there marginal differences between Rick and Sharon as they are co-authors, however Sharon beat Rick. Peter came last, and I am not sure why that is.
I also felt like the limited vocabulary of upgoer5 might harm concision as it may need more simple words to convey the same information, so I also made a metric based on the ratio of unique words, to text word length.
Let me know if either of you guys have any resource links for authors and content that you think should join the competition.
The same metric might do a good job at identifying texts of high quality that are mined with the original broad search queries: