More than 20 people from various companies in the educational technology area (EdTech) came together to join the discussion of human language learning technologies.
The discussion started with the topic of “the intersection of human language processing and machine learning”. During the event, more questions were asked and discussed from the perspectives of experts, learners and engineers, such as how could machine and technologies accelerate human language learning process? What kind of shortcomings do the existing approaches have for learners? How can we facilitate social interactions that are helpful to learning a language?
Attendees introduce themselves
During the discussion
Maijaliisa said, “Most of the language learning apps made available are on vocabulary and grammar practice, but it is limited, more things should be achieved by technologies such as real communication and creativity.”
Debbie added “There should be more engagement for learners, to spark their interest and it should be more relevant to their lifestyle.”
Martin said that it could be frustrating for learners when the learning process is approaching a plateau.
Giuseppe pointed out a key point,“Among building AI, NLP, deep learning and related tools for machine human language learning, the real problem is the complexity of the language. Is it the frequency, grammar or speech? Machine Language Learning should formulate curriculum that is relevant to you, your level of proficiency to best assess which type of topics are relevant to the users.”
Maijaliisa suggested that the applications could be equipped with the function of automated tagging, which enables corpus to be created quickly.
Tiffany was concerned that learners might have different needs for learning written language and spoken language through machine learning. Some students prefer face to face conversations as humans can listen and judge what's interesting and be encouraging.
Gregory said proficiency tests are helpful for all language learning processes such as the common European framework.
Martin pointed out the fact that the knowledge is squashed into students’ mind by textbook. Technologies should help learners to find their passion and help them learn languages as survival skills. The process of learning a language should stick to the basics, but because of the economies of scale, personalizing language learning is difficult.
Patrick brought up the problem that language learners need more chances to practice but everyone’s schedule is different. Due to needing time to practice, technologies could help to connect people from different cultures.
Ruting said, “The interaction between people is important for learning a new language, it should be humans communicating with humans instead of translations. Subtitles are also helpful.”
Conversely, Fred argued that subtitles are destructive to language learning.Immersion opportunities in text exchanges, such as translation integrated in WeChat, is preferable.
Mark talked about the dangers of relying on machine learning as it takes away people’s responsibilities to really understand meaning.
Cindy added, “Learning English should be personality dependent. For introverts, interacting with a machine can more opportunities to learners to practice spoken English.”
Tommy also agreed that it could be helpful to imitate conversations and make it customized and personalized.
We just sent you an email. Please click the link in the email to confirm your subscription!
OKSubscriptions powered by Strikingly