OLDaily

There's a new effort to map university degrees to workplace skills
Emily Bamforth, EdScoop, 2021/11/16

The National Student Clearinghouse is working with an AI company called AstrumU to map degrees and credentials to in-demand skills. The problem being addressed here is that "employers can’t easily find matches for their open positions who earned relevant experience outside of a traditional degree." The company is working on a "marketplace for recruiting and talent development based on quantifiable skills" and "identifying learning recommendations to fill skills gaps that allow an individual with a high probability of success to map to in-demand roles." Having a pipeline like this allows people to address the learning they need to find employment, and if this can be done efficiently, may allow them to develop a more broad-based learning strategy, if they wish.

Web: [Direct Link] [This Post]

Knowledge in Motion: How to Spot the Future in the Here and Now
Sharon Varney, De Gruyter, 2021/11/16

We see the future the way we see the past: not directly, but by reading the signs. That, essentially, is the premise of this article, and it offers generally good advice about how to go about it. This isn't the sort of futurism based on building scenarios, it's one based looking closely at the present to spot emerging patterns. "Conventional wisdom suggests that when large organisational change efforts fail, the problem is people’s resistance to change. Complexity science shows that it can be the complete opposite. Human beings are incredibly creative and adaptable."

Web: [Direct Link] [This Post]

A Proposed Architecture of an Intelligent System for Assessing the Student’s UML Class Diagram
Rhaydae Jebli, Jaber El Bouhdidi, Mohamed Yassin Chkouri, 2021/11/16

A Unified Modeling Language (UML) document is a set of standardized visual formats for representing objects and structure in software design. This paper (9 page PDF) offers a fairly detailed look at what would be involved to have an automated system parse and evaluate UML documents. It spends most of the time discussing the first part, describing various approaches as Simple API for XML (SAX) and Document Object Model (DOM). Of course, the hard part is the second part, where you actually compare UML models for assessment purposes. That said, I don't see anything in principle that would make this impossible, and it suggests to me that evaluating structured documents is a useful approach to AI-based assessments, especially if combined with good tools for creating them, like app.diagrams.net. Image: Stack Overflow.

Web: [Direct Link] [This Post]

Students are told not to use Wikipedia for research. But it’s a trustworthy source
Rachel Cunneen, Mathieu O'Neil, Academic Matters, 2021/11/16

Wikipedia is a fatalistic resource and I use it a lot myself, especially for concepts that are new to me (which happens pretty much every day, so as I said, I use it a lot). It's, um, 'reliable' (reliable in scare quotes) in the sense that it probably won't lead you astray, and is a good source of references for deeper reading, but it is quite slanted toward a particular perspective of what's important and what sources can be used; it could really use more diversity of authors and (especially) editors. It's also not error-free, in the sense that it often conflates two distinct concepts under one heading (they should remove entirely the 'forwarding' function). But even with these shortcomings, I would say it's not less biased or error prone than other sources at a similar level, and nothing comes close to it for range of coverage. I would tell my students "start with Wikipedia, but don't stop there."

Web: [Direct Link] [This Post]