In Consciousness and cognition
We analyzed similarities and differences between dream and waking reports in a collection by Kahan and Sullivan (2012). We compared word frequencies in the reports using the automatic text analysis software, Linguistic Inquiry and Word Count (LIWC). Social words were more frequent in dream than waking reports, support for Social Simulation Theory. Positive emotion words were less frequent in dream than waking reports, support for Threat Simulation theory. Amount of cognition was the same in dream and waking reports, i. e., the number of words in the summary category "Cognitive Processes" was not significantly different. But specific categories of cognitive words differed in frequency. We also applied a machine learning technique to dream research and built a support vector machine to achieve binary characteristics detection (here, whether a report is about waking or dream) based on LIWC word counts. Performance metrics for the support vector machine were high.
Zheng Xiaofang, Schweickert Richard
2022-Dec-01
Continuity hypothesis, Dream, Machine learning, Support vector machine, Text analysis
