Post by Niamh Walsh, Learnovate
In a memorable opening scene of the movie Robocop thirty years ago, a prototype security robot fails to reverse an instruction to terminate a target during a demonstration. The board watch in helpless horror as one of their members is eviscerated by a creation of their own making. I was reminded of this scene reading disturbing reports of a driverless car that killed a pedestrian in Arizona in spite of a test driver sitting behind the wheel.
Movies, books and comics often portray intelligent machines as suspicious, hostile and error prone. In this narrative, the lofty aspiration of the visionary who brought the robot into existence is contrasted with a crushing reality of disobedience and disaster. Whether to intentionally inspire or deliberately disturb, news headlines draw our attention only to the most dramatic instances of machine capabilities. The reality is that artificial intelligence (AI) and machine learning (ML) are already embedded in our workplaces, our schools and our homes.
You already enjoy the benefits of machine learning when Google Maps lets you know to expect heavy traffic. If it wasn’t for email spam filters driven by machine learning, your email would be inundated with offers of augmentation and beleaguered pleas to transfer money to foreign shores. You can thank (or blame) machine learning for the ranking of posts in your Facebook feed. Digital personal assistants, such as Amazon’s Alexa and Apple’s Siri, use machine learning to process your verbal request and figure out an appropriate response. When you use a chat application to access technical support, chances are you are speaking to an application who responds to your query with a list of recommendations from a knowledge base. Machine learning is already part of your daily life whether you realise it or not.
The terms artificial intelligence (AI) and machine learning (ML) are often used interchangeably but there is a difference. We use the term AI when computers do things that we usually expect only humans can do. Machine learning is an area of artificial intelligence in which computers learn from data rather than follow an explicit set of programming instructions. Humans provide data to machines and allow them to learn directly from the data. If the right conditions are in place, machine learning algorithms become more accurate over time as they process more and more data.
Technology is increasingly employed in learning whether at home, in work, at school or at university. Learning systems generate vast tracts of valuable data about learner journeys and preferences. As with other domains, analysing large data sets can be resource-intensive and require specialised skills that are not available to many small or medium-sized enterprises. Automating the analysis of learning data with machine learning can generate data-driven knowledge, which may improve the interactivity, personalisation and effectiveness of educational software. Adaptive learning is one area in which machine learning already plays a significant role. Employing machine learning algorithms can reduce administrative burden and overheads by automating manual tasks such as grading essays or content tagging. ML has also been used to notify school administrators when a student is at risk of dropping out, allowing them to intervene sooner.
We chose a use case to test if learning technology SMEs can benefit from off-the-shelf machine learning tools. We tested automating the tagging of learning content. Adding metadata to tag content is a manual, laborious process that requires a person to read each piece of content and decide how that content should be tagged. We wanted to find out if we could train a machine learning tool to tag learning content accurately by feeding it correctly tagged data. We developed a prototype that could apply what it learned from training to recommend tags for a fresh set of untagged data. If accurate, EdTech companies could use machine learning tools to speed up the tagging of content resulting in cost savings.
We started out asking whether EdTech SMEs should consider investing in machine learning. The benefits of AI and machine learning were once only available to a select few involved in scientific research, or later, to corporates with deep pockets and skilled resources. The challenge for small-to-medium-sized companies is deciding whether investing in machine learning will generate return on investment for their business.
Talk to us
If you want to find out how Learnovate can help you apply machine learning to solve the challenges your company faces, talk to Tom at tom.pollock@learnovatecentre.org or +353 1 896 4910.
Learnovate has published a report to help EdTech SMEs in Ireland understand the capabilities of off-the-shelf machine learning tools as machine learning becomes an increasingly important technology. The report will help EdTech companies who are wondering if machine learning is effective for small-to-medium-sized learning technology companies.
The report includes:
Learnovate’s industry-led collaborative research is undertaken for the collective benefit of the Irish learning technology industry. Funded by Enterprise Ireland, core research reports are available to Learnovate members. Complete the form below to request access to the report.
Stay up to date with the latest learning technology research, events and funding opportunities.
We will collect, use and protect your data according to our privacy policy.
Unit 28
Trinity Technology & Enterprise Campus
Pearse Street
Dublin 2
D02 N638
Ireland.
Tel. +353 1 8964910
Email: info@learnovatecentre.org
