Language Technologies for Learning
When people hear language technologies many immediately think about technologies for language learning. This is not correct. Language technologies are looking at interpreting human language and making sense of it. This is also different to the Semantic Web, a difference that I explained in this blog post a while ago.
Language technologies include, for example, sentiment analysis and opinion mining techniques which aim at categorising human expression into moods of expression. Natural language processing or latent semantic analysis (LSA) also try to "understand" human texts and compare them with each other to detect closeness or quality.
My own interest is based on a three year international research project that I managed. It focused on language technologies to support learners and tutors to make sense of large amounts of data, ways that helps them to save time or that lower the cognitive load. Among other things, language technologies can help analyse large multi-participant discussions (like chat and forums). With these analytics techniques, intertwined threads of chat utterances can be untangled and interaction between participants can be detected. This makes it easier to look at a long confused discussions and detect cross-links between individual participants.
Language technologies hold many promises for learning and knowledge construction. Similarly to Learning Analytics, they can help people to better understand their learning. What is best is that these are qualitative analyses that directly look at the language artefacts that learners have produced. In this way, they are unobtrusively observing learners while they learn and can provide feedback on demand at any time and independent of tutor availability.