The final project at my last role involved extracting seven years of learner data for 80,000 learners from an LMS that was being decommissioned. The shock of our vendor’s £150,000 quote to both extract and host 3TB of learner data led me to explore a robotic process automation (RPA) solution using a low cost browser plugin, UIVision. This article explores how I crafted and fine-tuned a suite of RPA scripts to automate the data extraction and reflects on the challenges faced and lessons learned.
About five years ago at Brightwave, we worked on a chatbot prototype for The Samaritans. The chatbot was a learning simulation tool, with the chatbot acting out the part of Bella, a bullied teenager. The training was for listening volunteers who were increasingly supporting people via chat channels rather than telephone calls. Bellabot offered a safe place to fail, to hone their responses on a text-only interface.
Bellabot was a completely scripted experience that only had a small range of conversational prompts about how she was being bullied and how that made her feel. As a volunteer, you had to listen to and chat with Bella. Despite the limitations, everyone who ‘talked to’ Bella was astounded at how emotionally engaged they became in the conversation. It was as if Bella was a real person, not a bot. Immersive is a word usually reserved for high tech, virtual reality simulations, but this low tech, text-based experience was one of the most immersive learning experiences I’d ever had. Now that Brightwave has been merged into Capita most of the original online blogs and presentations seem to have been taken offline which is a real loss, but the team ran quite a lot of webinars and conference sessions to talk about this at the time, it was a fantastic learning experience in how to use chatbots as a learning tool.
Fast forward to 2023 and you’d need to have been living under a rock to not know that chatbots have moved on somewhat since then. ChatGPT offers some amazing possibilities. One of the earliest and most intriguing uses of ChatGPT was when some child realised you could use it to generate text based MUD games, where you give the chatbot a scenario to play out and let a text based adventure unfold.
People have been trying to find educational uses for blockchainfor many years now. I try to carve out a little time every year to see if any research projects have matured into industry products but it’s never a productive search. There are always dedicated practitioners (mostly academics) running proof of concepts, and many more edtech commentators weighing in with blogs, ideas and conjecture about blockchain’s potential in education. But over the years things have progressed at such a glacial pace, I had pretty much concluded that blockchain in EdTech was possible just a solution that couldn’t find a problem.
I recently completed the Designing a Feminist Chatbot course on FutureLearn. As a starter course it has a lot going for it, covering areas such as different chatbot uses, how chatbots can become biased, user centred design (UCD) principles, persona creation, conversation design, storyboarding and prototyping. Having studied UCD at degree level and worked on countless software design projects in my career this could have become a bit boring but it’s always good to revisit the basics and is especially interesting to apply existing knowledge to new problems. There were some fascinating new areas to me such as conversation design and chatbot personality design, and the course drew heavily on the Google Conversation Design Process which is a great resource in itself and has some useful canvas-style templates for guiding the development of a chatbot.
Towards the end of 2020 our Head of Learning Design asked me to do a short presentation to her design team about where I thought the main disruptions to the Learning & Development technology market would come from in the year ahead. Usually we would look at startups, niche suppliers or parallel industries to identify potential disruptors. But if 2020 taught us anything, it was to think differently and look elsewhere for what could turn markets upside down!
Adaptive learning uses competence, behavioural and demographic data to tailor a digital learning experience around each learners unique needs. There’s a lot of hype around this area which might have you thinking its all about Artificial Intelligence (AI), but that’s not the case and there are two types of adaptive learning approaches: AI-based and Rule-based. Each will afford you different features, benefits and outcomes.
Reading Time:5minutesI spent two days last week at the Learning Technologies 2017 exhibition, working on the LEO stand (below). This annual event is split over two floors, with a paid conference upstairs and free exhibition downstairs. The stand was really busy for both days and the whole team came away absolutely exhausted, but I did manage to wander around the exhibition looking to see what the trends were this year and seeking out interesting new products.
There has been lots in the news this past year about social media bias and echo chambers, which started gaining prominence when algorithms started meddling in your news feed. The major web companies collect a huge amount of data about you and in doing so are building a detailed profile comprising demographic data, likes and purchases and other data that has been captured and purchased. As you ‘like’ posts and pages, so the algorithm delivers similar content back to you. Your friends like certain things, or ‘people like you’ like certain things, and the algorithm delivers more of that content to you too. You search for and purchase certain things, and you get delivered content related to that. Maybe you even give away valuable data via an innocuous-looking Facebook quiz, which is then sold to highest bidder and fed into yet more algorithms to target you with stuff you might ‘like’.
MoodleMoot UK and Ireland 2016 showed yet again that the Moodle ecosystem is in good health, with lots of new community members attending for the first time, plenty of old timers coming back, major institutions reaffirming their faith and Moodle HQ showing how the product itself is adapting to the future with new features and new sectors in its sights.
As learning analytics continues to rise up the agenda in the corporate learning & development (L&D) sector, one thing is becoming glaringly apparent: we should not expect a one-size-fits-all, off-the-shelf approach to learning analytics. This is a specialist discipline that cannot be bottled up into a single product. Sure, there are products such as Knewton, a Product as a Service platform used to power other peoples’ tools. There are also LMS bolt-ons like Desire2Learn Insights or Blackboard Analytics but even they are not sold as off-the-shelf products, for example the Blackboard team “tailors each solution to your unique institutional profile”. There are just far too many organisational factors at play for an L&D practitioner to be able to implement a learning analytics programme using an off-the-shelf tool.
