Leveraging Routine Behavior and Contextually-Filtered Features for Depression Detection among College Students

The rate of depression in college students is rising, which is known to increase suicide risk, lower academic performance and double the likelihood of dropping out of school. Existing work on finding relationships between passively sensed behavior and depression, as well as detecting depression, mainly derives relevant unimodal features from a single sensor. However, co-occurrence of values in multiple sensors may provide better features, because such features can describe behavior in context. We present a new method to extract contextually filtered features from passively collected, time-series mobile data via association rule mining. After calculating traditional unimodal features from the data, we extract rules that relate unimodal features to each other using association rule mining. We extract rules from each class separately (e.g., depression vs. non-depression). We introduce a new metric to select a subset of rules that distinguish between the two classes. From these rules, which capture the relationship between multiple unimodal features, we automatically extract contextually filtered features. These features are then fed into a traditional machine learning pipeline to detect the class of interest (in our case, depression), defined by whether a student has a high BDI-II score at the end of the semester. The behavior rules generated by our methods are highly interpretable representations of differences between classes. Our best model uses contextually-filtered features to significantly outperform a standard model that uses only unimodal features, by an average of 9.7% across a variety of metrics. We further verified the generalizability of our approach on a second dataset, and achieved very similar results.

[1]  Ramakrishnan Srikant,et al.  Mining sequential patterns , 1995, Proceedings of the Eleventh International Conference on Data Engineering.

[2]  Konrad Paul Kording,et al.  Mobile Phone Sensor Correlates of Depressive Symptom Severity in Daily-Life Behavior: An Exploratory Study , 2015, Journal of medical Internet research.

[3]  M. Thase,et al.  Depression, sleep, and antidepressants. , 1998, The Journal of clinical psychiatry.

[4]  Yoav Freund,et al.  A decision-theoretic generalization of on-line learning and an application to boosting , 1995, EuroCOLT.

[5]  Fanglin Chen,et al.  StudentLife: assessing mental health, academic performance and behavioral trends of college students using smartphones , 2014, UbiComp.

[6]  Emiliano Miluzzo,et al.  A survey of mobile phone sensing , 2010, IEEE Communications Magazine.

[7]  Fei-Fei Li,et al.  Visualizing and Understanding Recurrent Networks , 2015, ArXiv.

[8]  K. Dobson,et al.  Cognitive therapy of depression: pretreatment patient predictors of outcome. , 2002, Clinical psychology review.

[9]  F. Wahle,et al.  Mobile Sensing and Support for People With Depression: A Pilot Trial in the Wild , 2016, JMIR mHealth and uHealth.

[10]  Thomas Crady,et al.  College of the Overwhelmed: The Campus Mental Health Crisis and What to Do About It (review) , 2005 .

[11]  Jonathan W. Roberti,et al.  Factor structure, concurrent validity, and internal consistency of the beck depression inventory—second edition in a sample of college students , 2004, Depression and anxiety.

[12]  M S Magnusson,et al.  Discovering hidden time patterns in behavior: T-patterns and their detection , 2000, Behavior research methods, instruments, & computers : a journal of the Psychonomic Society, Inc.

[13]  Anind K. Dey,et al.  Leveraging Human Routine Models to Detect and Generate Human Behaviors , 2017, CHI.

[14]  Anind K. Dey,et al.  Modeling and Understanding Human Routine Behavior , 2016, CHI.

[15]  Laszlo Szathmary,et al.  Symbolic Data Mining Methods with the Coron Platform. (Méthodes symboliques de fouille de données avec la plate-forme Coron) , 2006 .

[16]  A. Kraskov,et al.  Estimating mutual information. , 2003, Physical review. E, Statistical, nonlinear, and soft matter physics.

[17]  R. Diler,et al.  College of the Overwhelmed: The Campus Mental Health Crisis and What to Do About It , 2005 .

[18]  Kaixiang Peng,et al.  Mining temporal association rules with frequent itemsets tree , 2018, Appl. Soft Comput..

[19]  Ramakrishnan Srikant,et al.  Fast Algorithms for Mining Association Rules in Large Databases , 1994, VLDB.

[20]  Philip S. Yu,et al.  Using a Hash-Based Method with Transaction Trimming for Mining Association Rules , 1997, IEEE Trans. Knowl. Data Eng..

[21]  Yoshihiko Suhara,et al.  DeepMood: Forecasting Depressed Mood Based on Self-Reported Histories via Recurrent Neural Networks , 2017, WWW.

[22]  Paul D. McNicholas,et al.  Standardising the lift of an association rule , 2008, Comput. Stat. Data Anal..

[23]  Szymon Jaroszewicz,et al.  A General Measure of Rule Interestingness , 2001, PKDD.

[24]  Saeed Moghaddam,et al.  MobileMiner: mining your frequent patterns on your phone , 2014, UbiComp.

[25]  Rakesh Agarwal,et al.  Fast Algorithms for Mining Association Rules , 1994, VLDB 1994.

[26]  Enhong Chen,et al.  An effective approach for mining mobile user habits , 2010, CIKM.

[27]  Jure Leskovec,et al.  Modeling Individual Cyclic Variation in Human Behavior , 2017, WWW.

[28]  Els Goetghebeur,et al.  Model evaluation based on the sampling distribution of estimated absolute prediction error , 2007 .

[29]  Matthew K Nock,et al.  Prevalence of and risk factors for suicide attempts versus suicide gestures: analysis of the National Comorbidity Survey. , 2006, Journal of abnormal psychology.

[30]  Toshi A Furukawa,et al.  Assessment of mood: guides for clinicians. , 2010, Journal of psychosomatic research.

[31]  J. Andrew Bagnell,et al.  Modeling Purposeful Adaptive Behavior with the Principle of Maximum Causal Entropy , 2010 .

[32]  Ashutosh Kumar Singh,et al.  Global, regional, and national incidence, prevalence, and years lived with disability for 310 diseases and injuries, 1990–2015: a systematic analysis for the Global Burden of Disease Study 2015 , 2016, Lancet.

[33]  Alexander Russell,et al.  Behavior vs. introspection: refining prediction of clinical depression via smartphone sensing data , 2016, 2016 IEEE Wireless Health (WH).

[34]  K. Demi̇rci̇,et al.  Relationship of Smartphone Use Severity with Sleep Quality, Depression, and Anxiety in University Students , 2015, Journal of behavioral addictions.

[35]  Guimei Liu,et al.  A Flexible Approach to Finding Representative Pattern Sets , 2014, IEEE Transactions on Knowledge and Data Engineering.

[36]  Howard J. Hamilton,et al.  Interestingness measures for data mining: A survey , 2006, CSUR.

[37]  Yuhong Yang,et al.  Cross-validation for selecting a model selection procedure , 2015 .

[38]  Jinmao Wei,et al.  Novel measurement for mining effective association rules , 2005, 2005 International Conference on Machine Learning and Cybernetics.

[39]  Hongxia Jin,et al.  RuleSelector: Selecting Conditional Action Rules from User Behavior Patterns , 2018, Proc. ACM Interact. Mob. Wearable Ubiquitous Technol..

[40]  Chloé Clavel,et al.  Using Temporal Association Rules for the Synthesis of Embodied Conversational Agents with a Specific Stance , 2016, IVA.

[41]  Deborah Estrin,et al.  Leveraging Multi-Modal Sensing for Mobile Health: A Case Review in Chronic Pain , 2016, IEEE Journal of Selected Topics in Signal Processing.

[42]  Laura E. Barnes,et al.  Using Mobile Sensing to Test Clinical Models of Depression, Social Anxiety, State Affect, and Social Isolation Among College Students , 2017, Journal of medical Internet research.

[43]  Anind K. Dey,et al.  Understanding and Using Context , 2001, Personal and Ubiquitous Computing.

[44]  Varun Jain,et al.  Depression Estimation Using Audiovisual Features and Fisher Vector Encoding , 2014, AVEC '14.

[45]  Gregory D. Abowd,et al.  Towards a Better Understanding of Context and Context-Awareness , 1999, HUC.

[46]  G. Brown,et al.  Social support, self-esteem and depression , 1986, Psychological Medicine.

[47]  Sriram Chellappan,et al.  Associating Internet Usage with Depressive Behavior Among College Students , 2012, IEEE Technology and Society Magazine.

[48]  Iqbal H. Sarker,et al.  Mining User Behavioral Rules from Smartphone Data through Association Analysis , 2018, PAKDD.

[49]  A. Doryab,et al.  Identifying Symptoms Using Technology , 2018 .

[50]  Oliver Brdiczka,et al.  Temporal task footprinting: identifying routine tasks by their temporal patterns , 2010, IUI '10.

[51]  Suman Nath,et al.  ACE: Exploiting Correlation for Energy-Efficient and Continuous Context Sensing , 2012, IEEE Transactions on Mobile Computing.

[52]  Denzil Ferreira,et al.  AWARE: Mobile Context Instrumentation Framework , 2015, Front. ICT.

[53]  David J. A. Dozois,et al.  A psychometric evaluation of the Beck Depression inventory-II , 1998 .

[54]  M. Madheswaran,et al.  Hybrid Medical Image Classification Using Association Rule Mining with Decision Tree Algorithm , 2010, ArXiv.

[55]  Jiawei Han,et al.  Extracting redundancy-aware top-k patterns , 2006, KDD '06.

[56]  Oliver Brdiczka,et al.  Using temporal patterns (t-patterns) to derive stress factors of routine tasks , 2009, CHI Extended Abstracts.

[57]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[58]  Yuan-Pang Wang,et al.  Response pattern of depressive symptoms among college students: What lies behind items of the Beck Depression Inventory-II? , 2018, Journal of affective disorders.

[59]  神田 信彦 Beck Depression Inventory-IIについての一考察 , 2004 .

[60]  Konrad Paul Kording,et al.  Distributed under Creative Commons Cc-by 4.0 the Relationship between Mobile Phone Location Sensor Data and Depressive Symptom Severity , 2022 .

[61]  Jian Pei,et al.  CMAR: accurate and efficient classification based on multiple class-association rules , 2001, Proceedings 2001 IEEE International Conference on Data Mining.

[62]  R A Steer,et al.  Mean Beck Depression Inventory–II Scores by Severity of Major Depressive Episode , 2001, Psychological reports.

[63]  Osmar R. Zaïane,et al.  Application of Data Mining Techniques for Medical Image Classification , 2001, MDM/KDD.

[64]  Alketa Hysenbegasi,et al.  The impact of depression on the academic productivity of university students. , 2005, The journal of mental health policy and economics.

[65]  R. Kessler,et al.  The epidemiology of depression across cultures. , 2013, Annual review of public health.

[66]  Jiawei Han,et al.  Summarizing itemset patterns: a profile-based approach , 2005, KDD '05.

[67]  Lars Vedel Kessing,et al.  Supporting disease insight through data analysis: refinements of the monarca self-assessment system , 2013, UbiComp.

[68]  Andrew T. Campbell,et al.  Next-generation psychiatric assessment: Using smartphone sensors to monitor behavior and mental health. , 2015, Psychiatric rehabilitation journal.

[69]  Jürgen Schmidhuber,et al.  Long Short-Term Memory , 1997, Neural Computation.

[70]  P. Duncan-Jones,et al.  Detecting anxiety and depression in general medical settings. , 1988, BMJ.

[71]  Dimitris Kanellopoulos,et al.  Association Rules Mining: A Recent Overview , 2006 .

[72]  Reda Alhajj,et al.  Effective Classification by Integrating Support Vector Machine and Association Rule Mining , 2006, IDEAL.

[73]  John Zimmerman,et al.  Detection of Behavior Change in People with Depression , 2014, AAAI Workshop: Modern Artificial Intelligence for Health Analytics.

[74]  M. Whisman,et al.  Normative Data on the Beck Depression Inventory--Second Edition (BDI-II) in College Students. , 2015, Journal of clinical psychology.

[75]  Daniel Eisenberg,et al.  Self-reported Barriers to Professional Help Seeking Among College Students at Elevated Risk for Suicide , 2013, Journal of American college health : J of ACH.

[76]  Vincent S. Tseng,et al.  ERMiner: Sequential Rule Mining Using Equivalence Classes , 2014, IDA.

[77]  Aristides Gionis,et al.  Mining Graph Evolution Rules , 2009, ECML/PKDD.

[78]  Wynne Hsu,et al.  Integrating Classification and Association Rule Mining , 1998, KDD.

[79]  Akito Monden,et al.  A hybrid faulty module prediction using association rule mining and logistic regression analysis , 2008, ESEM '08.

[80]  M. Kovács Cognitive therapy in depression. , 1980, The Journal of the American Academy of Psychoanalysis.

[81]  M. Von Korff,et al.  The Relationship Between Pain and Depression , 1996, British Journal of Psychiatry.

[82]  Rui Wang,et al.  Tracking Depression Dynamics in College Students Using Mobile Phone and Wearable Sensing , 2018, Proc. ACM Interact. Mob. Wearable Ubiquitous Technol..

[83]  Daniel Eisenberg,et al.  Help-Seeking and Access to Mental Health Care in a University Student Population , 2007, Medical care.

[84]  Jin-Mao Wei,et al.  Novel measurement for mining effective association rules , 2005 .

[85]  M. Silverman,et al.  Aspects of suicidal behavior, depression, and treatment in college students: results from the spring 2000 national college health assessment survey. , 2005, Suicide & life-threatening behavior.

[86]  Mohammad Hossein Namaki,et al.  Discovering Graph Temporal Association Rules , 2017, CIKM.

[87]  Jiawei Han,et al.  CPAR: Classification based on Predictive Association Rules , 2003, SDM.

[88]  Cecilia Mascolo,et al.  Mining users' significant driving routes with low-power sensors , 2014, SenSys.

[89]  N Sitaram,et al.  Sleep and depression. , 1987, Psychopathology.

[90]  Elgar Fleisch,et al.  Toward the Design of Evidence-Based Mental Health Information Systems for People With Depression: A Systematic Literature Review and Meta-Analysis , 2017, Journal of medical Internet research.

[91]  Katayoun Farrahi,et al.  Extracting Mobile Behavioral Patterns with the Distant N-Gram Topic Model , 2012, 2012 16th International Symposium on Wearable Computers.

[92]  Charles M Judd,et al.  Measurement Invariance of the Beck Depression Inventory–Second Edition (BDI-II) Across Gender, Race, and Ethnicity in College Students , 2013, Assessment.

[93]  Antonio Gomariz,et al.  SPMF: a Java open-source pattern mining library , 2014, J. Mach. Learn. Res..

[94]  D. Mohr,et al.  Harnessing Context Sensing to Develop a Mobile Intervention for Depression , 2011, Journal of medical Internet research.

[95]  A. McCallum,et al.  Topical N-Grams: Phrase and Topic Discovery, with an Application to Information Retrieval , 2007, Seventh IEEE International Conference on Data Mining (ICDM 2007).

[96]  G. Arbanas Diagnostic and Statistical Manual of Mental Disorders (DSM-5) , 2015 .

[97]  Mirco Musolesi,et al.  Trajectories of depression: unobtrusive monitoring of depressive states by means of smartphone mobility traces analysis , 2015, UbiComp.