Bipolar Disorder (BD) is a mental disorder characterized by intense mood swings, from depression to manic states. Individuals with BD are at a higher risk of suicide, but BD is often misdiagnosed as Major Depressive Disorder (MDD) due to shared symptoms, resulting in delays in appropriate treatment and increased suicide risk. While early intervention based on social media data has been explored to uncover latent BD risk, little attention has been paid to detecting BD from those misdiagnosed as MDD. Therefore, this study presents a novel approach for identifying BD risk in individuals initially misdiagnosed with MDD. A unique dataset, BD-Risk, is introduced, incorporating mental disorder types and BD mood levels verified by two clinical experts. The proposed multi-task learning for predicting BD risk and BD mood level outperforms the state-of-the-art baselines. Also, the proposed dynamic mood-aware attention can provide insights into the impact of BD mood on future risk, potentially aiding interventions for at-risk individuals.

Numerous datasets have been proposed to combat the spread of online hate. Despite these efforts, a majority of these resources are English-centric, primarily focusing on overt forms of hate. This research gap calls for developing high-quality corpora in diverse languages that also encapsulate more subtle hate expressions. This study introduces K-HATERS, a new corpus for hate speech detection in Korean, comprising approximately 192K news comments with target-specific offensiveness ratings. This resource is the largest offensive language corpus in Korean and is the first to offer target-specific ratings on a three-point Likert scale, enabling the detection of hate expressions in Korean across varying degrees of offensiveness. We conduct experiments showing the effectiveness of the proposed corpus, including a comparison with existing datasets. Additionally, to address potential noise and bias in human annotations, we explore a novel idea of adopting the Cognitive Reflection Test, which is widely used in social science for assessing an individual’s cognitive ability, as a proxy of labeling quality. Findings indicate that annotations from individuals with the lowest test scores tend to yield detection models that make biased predictions toward specific target groups and are less accurate. This study contributes to the NLP research on hate speech detection and resource construction. The code and dataset can be accessed at https://github.com/ssu-humane/K-HATERS.