Summary of the paper

Title Detecting Dementia from Repetition in Conversational Data of Regular Monitoring Service
Authors Kaoru Shinkawa and Keita Shimmei
Abstract Language dysfunctions are recognized as prominent signs of dementia, and previous computational studies have shown that measuring such dysfunctions can serve as a sensitive index of cognitive decline. These features of measuring language dysfunctions have been investigated in conversational data collected during neuropsychological tests but not in data collected during daily conversations. In this study, we used data obtained from a daily monitoring service for eight elderly people, including two who had been reported as having dementia, and investigated the features that characterize repetition in conversations on different days as well as single conversations on the same day. Through the analyses, we found that features for measuring repetition significantly increase for dementia patients in terms of topic and words. The results suggest that using the repetition features over the regular conversational data is a promising approach for detecting dementia sufferers.
Topics Monitoring Service, Linguistic Dysfunctions, Daily Conversation, Vocabulary Richness, Topic Similarity
Full paper Detecting Dementia from Repetition in Conversational Data of Regular Monitoring Service
Bibtex @InProceedings{SHINKAWA18.6,
  author = {Kaoru Shinkawa and Keita Shimmei},
  title = {Detecting Dementia from Repetition in Conversational Data of Regular Monitoring Service},
  booktitle = {Proceedings of the Eleventh International Conference on Language Resources and Evaluation (LREC 2018)},
  year = {2018},
  month = {may},
  date = {7-12},
  location = {Miyazaki, Japan},
  editor = {Dimitrios Kokkinakis},
  publisher = {European Language Resources Association (ELRA)},
  address = {Paris, France},
  isbn = {979-10-95546-26-9},
  language = {english}
  }
Powered by ELDA © 2018 ELDA/ELRA