Modeling Temporal Progression of Emotional Status in Mental Health Forum: A Recurrent Neural Net Approach

Patients turn to Online Health Commu- nities not only for information on spe- cific conditions but also for emotional sup- port. Previous research has indicated that the progression of emotional status can be studied through the linguistic patterns of an individual’s posts. We analyze a real- world dataset from the Mental Health sec- tion of healthboards.com. Estimated from the word usages in their posts, we find that the emotional progress across patients vary widely. We study the problem of predicting a pa- tient’s emotional status in the future from her past posts and we propose a Recur- rent Neural Network (RNN) based archi- tecture to address it. We find that the fu- ture emotional status can be predicted with reasonable accuracy given her historical posts and participation features. Our eval- uation results demonstrate the efficacy of our proposed architecture, by outperform- ing state-of-the-art approaches with over 0.13 reduction in Mean Absolute Error.

[1]  Jure Leskovec,et al.  No country for old members: user lifecycle and linguistic change in online communities , 2013, WWW.

[2]  Christos Faloutsos,et al.  Fast mining and forecasting of complex time-stamped events , 2012, KDD.

[3]  Yehuda Koren,et al.  Collaborative filtering with temporal dynamics , 2009, KDD.

[4]  Ted Pedersen,et al.  Why Do They Leave: Modeling Participation in Online Depression Forums , 2016, SocialNLP@EMNLP.

[5]  Maria Liakata,et al.  The language of mental health problems in social media , 2016, CLPsych@HLT-NAACL.

[6]  Eric Horvitz,et al.  Social media as a measurement tool of depression in populations , 2013, WebSci.

[7]  Mark Dredze,et al.  How Social Media Will Change Public Health , 2012, IEEE Intelligent Systems.

[8]  Steven C. Hillmer,et al.  An ARIMA-Model-Based Approach to Seasonal Adjustment , 1982 .

[9]  Minsu Park,et al.  Depressive Moods of Users Portrayed in Twitter , 2012 .

[10]  Charles Elkan,et al.  Learning to Diagnose with LSTM Recurrent Neural Networks , 2015, ICLR.

[11]  Le Zhang,et al.  Ensemble deep learning for regression and time series forecasting , 2014, 2014 IEEE Symposium on Computational Intelligence in Ensemble Learning (CIEL).

[12]  Kuldip K. Paliwal,et al.  Bidirectional recurrent neural networks , 1997, IEEE Trans. Signal Process..

[13]  Kunikazu Kobayashi,et al.  Time series forecasting using a deep belief network with restricted Boltzmann machines , 2014, Neurocomputing.

[14]  Mark Dredze,et al.  You Are What You Tweet: Analyzing Twitter for Public Health , 2011, ICWSM.

[15]  Christopher Heard,et al.  Wind Speed Prediction Using a Univariate ARIMA Model and a Multivariate NARX Model , 2016 .

[16]  Li Wei,et al.  Semi-supervised time series classification , 2006, KDD '06.

[17]  Jure Leskovec,et al.  Antisocial Behavior in Online Discussion Communities , 2015, ICWSM.

[18]  Amir F. Atiya,et al.  An Empirical Comparison of Machine Learning Models for Time Series Forecasting , 2010 .

[19]  Yoshua Bengio,et al.  On the Properties of Neural Machine Translation: Encoder–Decoder Approaches , 2014, SSST@EMNLP.

[20]  Eric Horvitz,et al.  Predicting Depression via Social Media , 2013, ICWSM.

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

[22]  Munmun De Choudhury,et al.  Characterizing Smoking and Drinking Abstinence from Social Media , 2015, HT.

[23]  Mark Dredze,et al.  Quantifying Mental Health Signals in Twitter , 2014, CLPsych@ACL.

[24]  Ted Pedersen,et al.  Analysis of Anxious Word Usage on Online Health Forums , 2016, Louhi@EMNLP.

[25]  Ping-Feng Pai,et al.  A hybrid ARIMA and support vector machines model in stock price forecasting , 2005 .