OLDaily

How to represent part-whole hierarchies in a neural network
Geoffrey Hinton, arXiv, 2021/04/23

There's a pretty good article about this in MIT Technology Review, but it's behind a paywall so I can't link to it (I have a full text version in my RSS feed reader; also see the comments here). The concept Hinton describes is called GLOM (derived from the slang "glom together"). The idea is simple but the tech is complex. Here's the idea: “Similarities of big vectors explain how neural networks do intuitive analogical reasoning.” A vector is an array of numbers that encodes information, for example, the xyz coordinates of a point. Any given perception can be represented as a really long vector - or as sets of multiple (and multidimensional) vectors. These subsets are similar to previously experienced vectors, allowing the neural net to extract parts from the whole. These are "islands of agreement". Why do I think this is a good idea? Because I had the same intuition in 1993 (note that I am in no way claiming to have discovered this; it's a very different thing to have an intuition and to flesh it out as a fully formed idea).

Web: [Direct Link] [This Post]

FAIR Competences for Higher Education
Yuri Demchenko, FAIRsFAIR, 2021/04/23

We want data that are findable, accessible, interoperable and reusable (FAIR). The FAIRsFAIR project is based on these concepts, proposing to "foster FAIR Data Practices in Europe" and beyond. This document is a set of competencies related to FAIR data management (and hence is directly relevant to a data literacy project I'm working on). It builds on and extends the the EDISON Data Science Framework (EDSF). Not yet officially endorsed, the proposed competence framework "is defined based on a recent job market analysis for the Data Steward and related professions." For more see the overall project website or this descriptions of the basics of FAIRsFAIR.

Web: [Direct Link] [This Post]

Teaching and Assessing Data Literacy
Cynthia Conn, et.al., Northern Arizona University, 2021/04/23

This is a resource guide (122 page PDF) to help support pre-service and in-service teachers. It's directly relevant to a current project I'm engaged in on defining and assessing data literacy. It begins with a good overview of data literacy frameworks, surveys a variety of resource sets for different aspects of data literacy, and references a number of data, statistical and graph literacy measures. Definitely one to keep on the virtual bookshelf.

Web: [Direct Link] [This Post]

What Happens When You Close the Door on Remote Proctoring? Moving Toward Authentic Assessments with a People-Centered Approach
Sarah Silverman, Autumm Caines, Christopher Casey, Belen Garcia de Hurtado, Jessica Riviere, Alfonso Sintjago, Carla Vecchiola, To Improve the Academy: A Journal of Educational Development, 2021/04/23

It's reasonable for institutions to discontinue the use of online proctoring services, but then what? This article looks at the impact of such a decision at the University of Michigan - Dearborn. Key is the provision of alternatives, for example, authentic assessments, "those that ask students to apply their knowledge on “intellectually worthy tasks and those should have the same competencies, or combinations of knowledge, skills, and attitudes, that [students] need to apply in the criterion situation in professional life." This needs to be supported with programming for instructors, say the authors, and in order to build trust should also be communicated effectively with students. (p.s. that's a lot of authors for a short and relatively introductory paper.)

Web: [Direct Link] [This Post]