Next panel: Mapping dynamic research ecosystem
What are ways to watch/monitor emerging technologies a la data? Measure/access what's important?
#CSPC2019
What are ways to watch/monitor emerging technologies a la data? Measure/access what's important?
#CSPC2019
The Naylor report, @cca_reports do point to areas of weakness or emerging topics. Also, internationally. What are they doing?
After that, what to do with that info?
Is there money to throw? Or policy actions (like inc. risk in peer review?)
#CSPC2019
After that, what to do with that info?
Is there money to throw? Or policy actions (like inc. risk in peer review?)
#CSPC2019
Next: Adam Bradley pointing to scholar eigenfactor to describe the network, citation rivers, and other data visualization of bibliometric data.
Move away from exploratory data analytics.... Context analysis key. Is that paper cited because it was wrong?
#CSPC2019
Move away from exploratory data analytics.... Context analysis key. Is that paper cited because it was wrong?
#CSPC2019
Love the poignant example of a dinosaur plot, deconstructed into eigenstates of various shapes, which all fit the same statistical measures...
Careful about the fitting! Radically different data sets with same metrics
#CSPC2019
Careful about the fitting! Radically different data sets with same metrics
#CSPC2019
Bibliometric studies, as described by @stefhaustein, need to compare 🍎 to 🍎, such as senior vs ECR research impact. New techniques are starting to allow for this.
Huge implications on funding evaluation and subsequent careers.
#CSPC2019
Huge implications on funding evaluation and subsequent careers.
#CSPC2019
.@stefhaustein talking on Twitter as an upcoming metric. Not used the same between disciplines (health sci vs math/physics... Vast differences). Hard to normalize.
#CSPC2019
#CSPC2019
Next Xiaodan Zhu on natural language processing (NLP) for taking on this issue.
If your computer can represent words (and their meaning) as a 3D vector, then off we go...
His lab models the meaning of words and reasoning in language. Bias? 🤨😬 Tricky stuff.
#CSPC2019
If your computer can represent words (and their meaning) as a 3D vector, then off we go...
His lab models the meaning of words and reasoning in language. Bias? 🤨😬 Tricky stuff.
#CSPC2019
Ex) all the reasons why you cite a paper.
Similarity between the paper you cite and paper you're writing.
#CSPC2019
Similarity between the paper you cite and paper you're writing.
#CSPC2019
What's the future of all this? Remember:
#CSPC2019
#CSPC2019
Data literacy will be a important for all disciplines.
Measuring academic impact?
Adam's example given: Twitter bots can be made to promote certain papers, skewing metrics... This can be done right now.
#CSPC2019
Measuring academic impact?
Adam's example given: Twitter bots can be made to promote certain papers, skewing metrics... This can be done right now.
#CSPC2019
Dr. Boston made a good point, need subject matter expertise + the data/analysis.
Shout out to Librarians. They have been doing this stuff for years... now add with comp sci, ML, and data to their expertise
#CSPC2019
Shout out to Librarians. They have been doing this stuff for years... now add with comp sci, ML, and data to their expertise
#CSPC2019
Nice point by Adam Bradley, machine learning categorizes stuff differently than humans. Tread carefully.
Another great question: distillation of years of research into short material useful for policy briefs?
#CSPC2019
Another great question: distillation of years of research into short material useful for policy briefs?
#CSPC2019
@threadreaderapp unroll pls
