I LET DEPRESSION AND ANXIETY DROWN ME…” IDENTIFYING FACTORS ASSOCIATEDWITH RESILIENCE BASED ON JOURNALING USING MACHINE LEARNING AND THEMATIC
Abstract : Over the years, there has been a global increase in the use of technology to deliver interventions for health and wellness, such as improving people’s mental health and resilience. An example of such technology is the Q-Life app which aims to improve people’s resilience to stress and adverse life events through various coping mechanisms, including journaling. Using a combination of sentiment analysis and thematic analysis methods, this paper presents the results of analyzing 6023 journal entries from 755 users. We uncover both positive and negative factors that are associated with resilience.
? This approach has been adopted by only a few existing research in mental health domain to the best of our knowledge.
? In our dataset, a total of 68247 reviews have user ratings (representing 77.4% of total reviews). Similar to the approach used by existing research we applied the criteria in to automatically annotate reviews. presents the number of reviews per sentiment polarity after annotation.
? Our approach aligns with existing research which defines sentiment classification as a ‘‘two-class classification problem, positive and negative’’ and also as a technique that ‘‘determines whether a document or sentence is opinionated, and if so whether it carries a positive or negative opinion’’
? Complement NB addresses skewed training data problem by estimating model weights from the complement of each class thereby making it well suited for imbalanced datasets.
? Logist Reg is a linear model for obtaining probabilities describing possible outcomes (binary, multinomial, or ordinal) using a logistic function that minimizes the loss on a dataset
? Research also shows that a high level of mindfulness and relaxation will significantly increase resilience and reduce the risk of developing mental health problems
? Future work will extend this study by applying machine learning techniques to automatically detect thematic problems of individuals based on their journal entries and then propose appropriate interventions.
? To address the negative factors discussed above, we propose the following design recommendations based on the positive themes in (see Appendix section) which are factors that contribute to the effectiveness of mental health apps.
? HACP, while very welcome, is in ways reminiscent of the early days of the Web, when many “similar” quality benchmarking tools and codes of conduct for information publishers were proposed to appraise and rate online medical and health information.
? The health sector has used this approach to provide real-time health services and early intervention through web and mobile applications
? Journaling in healthcare has been used primarily for counseling and psychotherapy to provide self-help interventions and promote personal reflection
? Sentiment analysis is used to analyze people’s opinions, sentiments, evaluations, attitudes, and emotions from written language
? we utilized and compared multiple classifiers using lexicon-based and machine learning techniques to determine the best performing classifier overall, which is then used in predicting the sentiment polarity of unclassified journal entries.
|