And so, in the spirit of working in the open and embracing social learning, we think it’s time that we shared some of our process and insights as we build the learning technology tools of the future.
Starting Spring 2017, each week we’ll take a topic which is key to our research, like text analytics, machine learning and learning analytics, and start with an industry or business viewpoint of key issue or topic – and then take a look at current research that is informing design and practice.
We’ll send out a weekly ‘Research Digest’ to summarise and highlight some of the main reading we’ve been doing over the past 7 days, and also post a monthly ‘Abstracts’ article here on the blog in which we’ll also share the details of how we’re applying our new knowledge.
To give you a feel for what these new resources will look like, here’s what we’ve been reading this week…
Research Digest: Artificial Intelligence
In this week’s digest we dive headlong into the promises of Machine Learning, the difficulty of Text Analytics and the potential influence of visualisations on learning outcomes…
How Artificial Intelligence Will Change Everything (Wall Street Journal, 7 March 2017)
This article, while somewhat provocatively titled, is a very practically focused interview with Andrew Ng of Baidu and Neil Jacobstein of Singularity University. They see the adoption of AI as something very real and active in the business world, and survey data seems to back this up.
Bringing Order to Chaos in OLX Discussion Forums with Content-Related Thread Identification (Wise, et al – p. 188-197.)
A key driver for us at Learning Pool (formerly HT2 Labs) is to better understand what our learners are saying in social learning environments. This paper looks at one of the more fundamental problems; can we understand the direction a conversation is heading, without knowing about the specific knowledge domain people are talking about?
Identifying content related threads is a basic building block that is key to supporting a variety of strategic and analytics goals… this approach seems promising for building base models that may be usable without much customisation across a variety of courses.
Relevance Based Language Models (Lavrenko and Croft)
A serious challenge to meaningful Text Analysis is acquiring enough training data to make analysis meaningful. Typically, we spend days training our models with painstakingly hand-marked sets of data, to then test automated models against. But this paper proposes taking a different approach to the normal, Bayesian-Probability style of text analysis, which could save a lot of time and energy…
This approach points to promising analogies for use in document classification in text analytics. The possibility of using our potential data queries as a basis of document analysis is an exciting one.
The Role of Achievement Goal Orientations When Studying Effect of Learning Analytics Visualizations (Beheshitha, et al.)
A key part of social learning is self-efficacy; that is, how you judge your own performance. A lot of traditional eLearning takes place in isolation, robbing people of the chance to see how they are doing compared to others. This study shows the potential impact of showing benchmarking visualisations to learners.
Given the demand for effective dashboard design, and for individualised feedback, understanding the impact of learning analytics on student performance is essential for effective design. This study underscores the need to know the learners and the strategic goals for a learning activity in order to design dashboards that encourage mastery and performance.
Stay Briefed, Get Involved
If you’ve made it to the end of this post, then you’re probably the type of person that these resources are aimed at and who’ll find them of most use, so here’s how you can stay briefed and get involved:
And if you’re not already doing so, you can also follow us @learningpool on Twitter for all the latest news, views, blogs and reviews from The Lab.
The first edition of the Research Digest will be published at the end of March, so if there’s anything in particular that you think either we, or the broader learning community should be thinking about right now, do let us know!
Get started by telling us what you need and one of our team will be in touch very soon.
+44 207 101 9383
US +1 857 284 1420
+44 345 074 4114*
US +1 844 238 5577
* call charges vary depending on your provider