Long version: Thanks to Luciano Gerber, Lutz Büch, and Gerlof Bouma for their responses and references! Indeed, there are a number of potential terms. I followed up on their messages with some Google Scholar searches. These mostly confirmed something that Gerlof observed: we don't use them enough! Lacking quantitative data (other than that most searches yielded a result of 0 hits), and examining the references that I received, I can summarize the qualitative facts like this:
- Options include either "line" (e.g. "isoline" or "contour line") or "curve" (e.g. "isometric curve") - Options include either "iso-" (e.g. "isoline" or "iso-fscore") or "contour" (e.g. "contour line")
...in their various combinations. Given that:
- "Curve" is a better descriptor of what the range of F-measure is than is "line" - "isometric" captures the semantics of what an F-measure is better than does "contour"
...*isometric curve* appears to be the best term for our purposes.
Thanks again to Luciano, Lutz, and Gerlof!
Kevin *Original question follows...*
---------- Forwarded message --------- From: Kevin B. Cohen <kevin.cohen at gmail.com> Date: Thu, Jun 24, 2021 at 11:09 AM Subject: Looking for the right term for... To: Corpora List <corpora at uib.no>
I am looking for the proper term for a kind of line on a scatter plot that displays recall, precision, and F-measure.
Below there is an example of a plot with the lines in question, from this blog post <http://gromgull.net/blog/2009/06/plotting-precision-and-recall-values/> (for which I thank the unnamed author). You will see that it has precision on the x-axis, recall on the y-axis, and then lines that show the boundaries of regions where the F1 has a particular value corresponding to the values of R and of P on the x- and y-axes.
It's those lines whose name I am looking for. In meteorology they indicate air pressure, and are called isobars. In dialectology, they go on maps, where they mark geographical boundaries between lexical items, and are called isoglosses. Wikipedia has an article on isosurfaces <https://en.wikipedia.org/wiki/Isosurface>, and that might be the right word, but I would not swear to it. Also, if there is a common term in NLP, I would love to use it.
Kevin Cohen -- Kevin Bretonnel Cohen, PhD Director, Biomedical Text Mining Group Computational Bioscience Program, U. Colorado School of Medicine D'Alembert Chair in Natural Language Processing for the Biomedical Domain (Emeritus), LIMSI, CNRS, Université Paris-Saclay 303-916-2417 http://compbio.ucdenver.edu/Hunter_lab/Cohen
-- Kevin Bretonnel Cohen, PhD Director, Biomedical Text Mining Group Computational Bioscience Program, U. Colorado School of Medicine D'Alembert Chair in Natural Language Processing for the Biomedical Domain (Emeritus), LIMSI, CNRS, Université Paris-Saclay 303-916-2417 http://compbio.ucdenver.edu/Hunter_lab/Cohen -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: text/html Size: 7247 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20210730/5fd03094/attachment.txt> -------------- next part -------------- A non-text attachment was scrubbed... Name: 3653326007_bd5f1a4a0e_o.png Type: image/png Size: 31482 bytes Desc: not available URL: <https://mailman.uib.no/public/corpora/attachments/20210730/5fd03094/attachment.png>