...

How we can improve sleep quality by finding the cross-section analysis of features

HowExtracting psycholinguistic features and finding correlations with SleepHealth collected features

each participant in this study is requested to complete two journals per day, one in the early morning and one before sleep. PsyLaP pipeline is used to extract psycholinguistic features from the provided free text.

In addition to the journals, individuals fill several self-assessed tasks in two different frequencies: daily and once. These tasks collect information about personalized sleep-related issues and personal information.

Once both feature sets are gathered, cross-section analysis is performed to indicate the association between journal and self-assessed features. Experiments are done in two different approaches, single association finding for each person, and longitudinal analysis of each individual over time.

So far, some single associations are found and are included and submitted to SBM2022.

Further reading

SBM2022 abstract

Real-world longitudinal data collected from the SleepHealth mobile app study

Additional information

Nature Paper:

NLP features are currently extracted and ready for further analysis.

Available features are:

Polarity features

Readability features

Topic Modeling

LIWChttps://www.nature.com/articles/s41597-020-00753-2

Resource Stack