OLDaily

Closer to the Machine
Sven Bluemmel, Office of the Victorian Information Commissioner, 2019/08/29

This book (155 page PDF) was published "to increase the Victorian public sector’s understanding of AI technologies that have the potential to impact our lives, and to assist those who implement AI systems to appreciate the technical, social, and legal aspects." It's not one single work but rather seven individual essays by seven authors on topics ranging from defining AI to where AI can go wrong to issues in regulating AI. Add this one to your library of ethics-in-AI resources.

Web: [Direct Link] [This Post]

Chatbot and Related Research Paper Notes with Images
Richard Csaky, GitHub, 2019/08/29

This page lists dozens and dozens of "papers related to chatbot models in chronological order spanning about 5 years from 2014... a starting point to get a hang of what the paper is about, and to mention main concepts with the help of pictures." The notes are really insightful, and some great papers are listed, including papers that ask hard questions about what constitutes things like common sense, dialogue, negotiation, engagement or cooperation (for example). Related (and included in this list): Neural Approaches to Conversational AI, "a good in-depth introduction to the many aspects of conversational AI as a field."

Web: [Direct Link] [This Post]

IMS Compliant Ontological Learner Model for Adaptive E-Learning Environments
Othmane Zine, Aziz Derouich, Abdennebi Talbi, International Journal of Emerging Technologies in Learning, 2019/08/29

The first sentence of this paper (23 page PDF) makes me want to reframe the learning styles debate. Here's the sentence: "It has been proven that adopting the 'one size fits one' approach has better learning outcomes than the 'one size fits all' one." The paper cites several references to say "an accurate definition the learner’s characteristics influences and increases considerably the capability and efficiency of learning activities." Well, this is learning styles, isn't it? Maybe not a four-term categorization like Kolb's or Fleming's models, but still the same concept. If the learning styles sceptics are right, then it should be true that "one size fits all". This paper adopts exactly the opposite posture, and indeed, so does most literature based on learner models, customization and personalization.

Web: [Direct Link] [This Post]

Knowledge Management and Reuse in Virtual Learning Communities
Houda Sekkal, Naila Amrous, Samir Bennani, International Journal of Emerging Technologies in Learning, 2019/08/29

The idea of capturing knowledge from informal communication has been around for decades - see for example this proposal from 2003 called Semex - but there are different ways to capture this knowledge and different kinds of knowledge are captured. This is a relatively short paper (17 page PDF) that looks at those options. It's based on an analysis of several knowledge-capture papers using something called the 'Markus Knowledge Reuse Process',  which is in reality a simple four-step model consisting of capturing, packaging, distribuition, and reuse. It occurs to me that these processes actually create knowledge (or at the very least, information) rather than replicating what is already known, so what follows doesn't really qualify as 'reuse' but rather 'application'.

Web: [Direct Link] [This Post]