OLDaily

When #edutwitter celebrities say stupid things
Bill Fitzgerald, Twitter, 2019/08/07

"If you choose to respond, screenshot the tweet, and respond to your screenshot." So says Bill Fitzgerald in this Twitter thread. And it raises obvious questions. Like: there are #edutwitter celebrities? What do they look like? The very next Twitter thread gives us an answer, with some observations by Audrey Watters and Akil Bello introducing us to the phrase “toxic positivity”. Fitzgerald also points us to this thread from Jennifer Binis, who asks "about who gets elevated to the status of EduCelebrity." Image: me taking Fitzgerald's advice. More: @EduCelebrity Twitter profile.

Web: [Direct Link] [This Post]

Results of OA article data collection from OASPA members
Heather Morrison, Global Open Access List (GOAL), 2019/08/07

This is discussion of data obtained for this study, which was reported in early July. Commenting on the results, Heather Morrison writes, "I recommend against the use of licenses allowing blanket commercial re-use to authors, journals, OA advocates and policy-makers." Why? Because the commercial indexing services index both open and closed-access publications, while non-commercial services are frozen out from commercial sources of data. This, she writes, results in an "increase in monopoly power for Elsevier: anyone can use the CC licensed material to create a competitor to Scopus, however only Elsevier can use their copyrighted work. CC-BY reduces the likelihood of successful competition." We've seen this same problem in open source software as well. There's no reciprocity; commercial enterprises take from the open, but never contribute back, and eventually undermine the open entirely.

Web: [Direct Link] [This Post]

Neural Networks: Feedforward and Backpropagation Explained & Optimization
Casper Bøgeskov Hansen, Machine Learning From Scratch, 2019/08/07

I really appreciate the effort in this post, though the lack of facility with the language makes it impossible to understand in places. Skip the overview. Start reading at the section titled "What is a neural network?" It might still leave you confused, but I find that the effort the author takes to make things clear outweighs the difficulty his use of language might otherwise cause. Even more to the point, this is a really good example of someone learning by sharing what he has learned by writing it out as clearly as possibe. It's a practice to be emulated!

Web: [Direct Link] [This Post]

Global labor flow network reveals the hierarchical organization and dynamics of geo-industrial clusters
Jaehyuk Park, et.al., Nature Communications, 2019/08/07

This is some interesting work that has been summarized in a number of places (Science Daily, Phys.org, Reddit (auto-summarized), Science World Report). In essence, the authors "construct a labor flow network of over 4 million firms across the world' using employment data from LinkedIn. This is used to draw conclusions about the nature of industrial clusters and of economic development. Successful geo-clusters "exhibit a stronger association between the influx of educated workers and financial performance.... Workers tend to change their jobs between geographically close firms with similar skill requirements. This tendency leads to knowledge spillover and innovation, serving as a prominent feedback mechanism in the formation of geo-industrial clusters."

Web: [Direct Link] [This Post]

Jacques Derrida
Leonard Lawlor, Stanford Encyclopedia of Philosophy, 2019/08/07

This is an extensively revised article on one of the more interesting (and difficult) philosophers of the 20th century, Jacques Derrida. I won't pretend to be able to speak with any authority on his work. But here's the ten-cent perspective of at least some of it: there are no simple irreducible concepts. Every concept (every work, every perception) contains both itself and the negation of itself inherently in its presentation. For example, "what is happening right now is also not different from every other now I have ever experienced. At the same time, the present experience is an event and it is not an event because it is repeatable." Or for example, "for a decision to be just, not only must a judge follow a rule but also he or she must 're-institute' it, in a new judgment. Thus a decision aiming at justice (a free decision) is both regulated and unregulated." What's important here (to me) is that you can't separate these different aspects of the concept; they are one and the same thing.

Web: [Direct Link] [This Post]

Articles Summaries that “Spoil” the Paper to Save Reader Time
Lisa Janicke Hinchliffe, The Scholarly Kitchen, 2019/08/07

Making an article summarizer has been a long-standing ambition of technologists. I remember when I first started at NRC in 2001 being told about Copernic, our summarizer work-in-progress. It worked pretty well and all it lacked as an API to integrate it into other software. Well, it was the early 2000s. Today for some reason now we read about Paper Digest, an application that does essentially the same thing. You can try it out at their web interface. Here's a sample of a paper that has been summarized. It depends a lot on section headers and seems to extract what it feels are relevant sentences. There are many more summarizers out there on the internet; here are some: Free Summarizer, Resoomer, Smmry, Summarizing, Text Summarizer, Autosummarizer, Tools4noobs (which includes a WordPress widget), appZaza, SummarizeThis, and more. Not that I'm worried about any of these taking my job. They don't do the background research, and they don't have the attitude.

Web: [Direct Link] [This Post]

The student loan crisis won’t disappear unless we tackle the root cause
Alana Dunagan, Christensen Institute, 2019/08/07

Somewhere along the line the Christensen Institute turned from advancing a realitively reasonable set of propositions about innovation to some sort of ideologically-based commentary on education and other industries. Here's the latest: "there’s one solution that not only offers potential students a measure of security against taking on loans they can’t pay off, but also protects against spending time on an education that doesn’t align with workforce needs: outcomes-based funding." Seriously? Has outcomes-based pricing made lawyers and law any cheaper? Has it pushed down the price of professional sports? Is there no way outcomes-based pricing can be manipulated to keep people paying high prices?

Web: [Direct Link] [This Post]