figshare
Browse
1/1
3 files

Conversation in elderly people

Version 3 2020-11-15, 17:02
Version 2 2020-11-12, 11:09
Version 1 2020-11-12, 11:07
dataset
posted on 2020-11-15, 17:02 authored by Masato AbeMasato Abe
[data.txt]
We recorded conversations of groups of four or five with 65 older Japanese people between June and September 2018. There were 16 groups and they were fixed over 14 sessions. In addition, as a condition, eight groups were required to speak freely and the remaining eight groups were required to make a presentation. Each session took place once a week for approximately 30 minutes. We decomposed the raw conversation into words and lemmatized the words. All words have been replaced with numbers to exclude personal information. The word column in the data represents words; the id column represents participants from 1 to 65; session represents session numbers from 1 to 14; group represents 16 groups from A to P; condition represents either free conversation or presentation; and the word column represents the number of participants. The words are arranged in the order of time.

[data_heaps.txt]
This file contains the number of unique words when the number of words increases. Token and type means the number of words and the number of unique words, respectively. This files was transformed from data.txt.

[cognitive_score.csv]
This file contains cognitive scores and other characteristics for 65 participants.

History