Deep Hierarchical Ensemble Model for Suicide Detection on Imbalanced Social Media Data

Attention-Enabled Ensemble Deep Learning Models and Their Validation for Depression Detection: A Domain Adoption Paradigm

Depression is increasingly prevalent, leading to higher suicide risk. Depression detection and sentimental analysis of text inputs in cross-domain frameworks are challenging. Solo deep learning (SDL) and ensemble deep learning (EDL) models are not robust enough. Recently, attention mechanisms have been introduced in SDL. We hypothesize that attention-enabled EDL (aeEDL) architectures are superior compared to attention-not-enabled SDL (aneSDL) or aeSDL models. We designed EDL-based architectures with attention blocks to build eleven kinds of SDL model and five kinds of EDL model on four domain-specific datasets. We scientifically validated our models by comparing "seen" and "unseen" paradigms (SUP). We benchmarked our results against the SemEval (2016) sentimental dataset and established reliability tests. The mean increase in accuracy for EDL over their corresponding SDL components was 4.49%. Regarding the effect of attention block, the increase in the mean accuracy (AUC) of aeSDL over aneSDL was 2.58% (1.73%), and the increase in the mean accuracy (AUC) of aeEDL over aneEDL was 2.76% (2.80%). When comparing EDL vs. SDL for non-attention and attention, the mean aneEDL was greater than aneSDL by 4.82% (3.71%), and the mean aeEDL was greater than aeSDL by 5.06% (4.81%). For the benchmarking dataset (SemEval), the best-performing aeEDL model (ALBERT+BERT-BiLSTM) was superior to the best aeSDL (BERT-BiLSTM) model by 3.86%. Our scientific validation and robust design showed a difference of only 2.7% in SUP, thereby meeting the regulatory constraints. We validated all our hypotheses and further demonstrated that aeEDL is a very effective and generalized method for detecting symptoms of depression in cross-domain settings.

Detecting Suicidal Ideation in Social Media: An Ensemble Method Based on Feature Fusion

Suicide has become a serious problem, and how to prevent suicide has become a very important research topic. Social media provides an ideal platform for monitoring suicidal ideation. This paper presents an integrated model for multidimensional information fusion. By integrating the best classification models determined by single and multiple features, different feature information is combined to better identify suicidal posts in online social media. This approach was assessed with a dataset formed from 40,222 posts annotated by Weibo. By integrating the best classification model of single features and multidimensional features, the proposed model ((BSC + RFS)-fs, WEC-fs) achieved 80.61% accuracy and a 79.20% F1-score. Other representative text information representation methods and demographic factors related to suicide may also be important predictors of suicide, which were not considered in this study. To the best of our knowledge, this is the good try that feature combination and ensemble algorithms have been fused to detect user-generated content with suicidal ideation. The findings suggest that feature combinations do not always work well, and that an appropriate combination strategy can make classification models work better. There are differences in the information contained in different functional carriers, and a targeted choice classification model may improve the detection rate of suicidal ideation.