Detecting dementia in Mandarin Chinese using transfer learning from a parallel corpus
This addresses the scarcity of data for dementia detection in non-English languages, though it appears incremental as it adapts existing transfer learning approaches to a new domain.
The paper tackled the problem of detecting Alzheimer's disease in Mandarin Chinese speech by learning a correspondence between lexicosyntactic features across languages using a parallel corpus, and demonstrated that their method outperformed baselines.
Machine learning has shown promise for automatic detection of Alzheimer's disease (AD) through speech; however, efforts are hampered by a scarcity of data, especially in languages other than English. We propose a method to learn a correspondence between independently engineered lexicosyntactic features in two languages, using a large parallel corpus of out-of-domain movie dialogue data. We apply it to dementia detection in Mandarin Chinese, and demonstrate that our method outperforms both unilingual and machine translation-based baselines. This appears to be the first study that transfers feature domains in detecting cognitive decline.