Televic shares its latest findings on AI in Digital Assessment (AIDA)
The ‘International e-Assessment Conference & Awards’ is one of the world’s leading conferences focusing on digital assessment. At this year’s edition, which was held in London, Televic shared its lessons learned from three use cases of converting AI research results into usable product features.
The outcomes were presented by Professor Thomas Demeester, from the IDLab research group at the University of Ghent, and Filip Vanlerberghe, Product Manager at Televic Education.
A world of possibilities
While AI offers lots of possibilities, the presenters argued that a well-functioning algorithm is not automatically usable for end users in their day-to-day tasks. Some key takeaways:
- The amount and quality of available data defines the accuracy of your algorithm.
- Fewer data can still result in very specific and narrow implementations of AI.
- Human post-editing of the results of AI-powered content creation significantly increases the quality of the results.
- Technology companies can make a difference not only with algorithms, but even more so with a strong focus on usability, keeping the technicality away from the user.