ABSTRACT Integrating deep learning with interpretable machine learning methods offers significant benefits for predicting mental health risks. This research introduces a hybrid architecture MH‐XAI, aimed at precisely predicting mental health while providing transparent, feature‐level explanations. MH‐XAI integrates a multi‐scale 1D Convolutional Neural Network (CNN), channel‐wise attention mechanisms, an ensemble of (XGBoost) classifiers, and SHAP (SHapley Additive exPlanations) to achieve both local and global interpretability. The model is trained on a comprehensive dataset derived from Open Sourcing Mental Illness (OSMI) Mental Health in Tech surveys conducted between 2016 and 2023. This dataset encompasses a variety of demographic, psychological, and occupational characteristics. The input features are restructured into a format that facilitates deep convolutional learning. The CNN feature extractor utilizes parallel convolutional layers with three different kernel sizes, allowing the model to capture both short‐range and long‐range dependencies in tabular data. The multi‐scale representations are subsequently enhanced using a channel‐wise attention process. The acquired features are transmitted to an ensemble of XGBoost classifiers for enhanced prediction accuracy. MH‐XAI attains a test accuracy of 91.54%, F1‐score of 92%, precision of 92%, and recall of 91%, surpassing standalone CNN and XGBoost. SHAP elucidates the model's predictions by quantifying the contributions of each feature. Results underscore critical factors such as past mental health history, workplace culture, current mental health, and family history that affect outcomes. MH‐XAI provides a precise, scalable, and comprehensible solution for the early detection of mental health in workplace environments.

ABSTRACT Regret is a complex negative emotion often associated with feelings of remorse, self‐blame, and disappointment regarding past actions or decisions. It plays a significant role in various business and decision‐making contexts and also has an impact on the health of individuals. This work aims at the detection of regret‐one of the most important emotion. Existing research on regret detection has been predominantly limited to English content. It is observed that people find it easier to communicate their feelings effectively in their native or code‐mixed languages. However, there is no work focusing on the detection of regret from text written in these languages. To address this gap, this paper first presents a novel dataset in Hindi using posts/comments written in Hindi or Hindi Roman script from multiple sources, incorporating both manual and automated annotation techniques to enhance the quality and consistency of data labeling. Then, it proposes a multi‐class regret detection framework to detect regret and classify its domain. The proposed framework HRegBERT‐CNN integrates a fine‐tuned BERT(regret) model for Hindi with CNN using N‐gram word embeddings, enabling it to capture local contextual features and complex patterns in the text effectively. Experimental results show that the HRegBERT‐CNN model outperforms state‐of‐the‐art models on the Hindi regret dataset by at least 3% and 5% for regret detection and domain identification tasks, respectively, in terms of macro F1‐score.

The volume of hate speeches and the number of user-generated materials are steadily rising, notably on social media networks. This trend can be seen across the internet. Therefore, it is necessary to recognize this kind of offensive content and remove it to maintain the cleanness of the platform. In spite of the fact that pertinent research was conducted separately for detecting hate speeches and social media code-mixed texts, purpose of this work is to identify hate speeches from social media code-mixed text. In the presented research, experiments for detecting hate speech using Bidirectional Encoder Representations from Transformers (BERT) architecture are carried out with the accessible code-mixed dataset, and optimization is carried out using the Particle Swarm Optimization algorithm (PSO). The results of our proposed methodology showcased notable improvements in the performance over standard BERT model while achieving an accuracy of 95.37% and an F1-score of 95.30%. Further, testing on an additional dataset confirms the generalizability of our approach, maintaining a high accuracy of 94.5%. These findings validate the effectiveness of PSO-driven optimization for hate speech detection in low-resource, code-mixed linguistic settings. To offer a comprehensive assessment of our approach, a comparative analysis with several classifiers, including Naive Bayes, Random Forest, XG Boost, CatBoost, KNN, Decision Tree, Adaboost, SVM, and LSTM has also been done.

SUMMARY Due to the tremendous growth of video editing software, it has become extremely simple to introduce malicious content by manipulating multimedia data. This may include modification of videos either by adding or deleting selective frames with malicious intentions. Hence, it is essential to find the forged frames of the videos by introducing efficient and reliable video forensic methods. This article presents an automatic intra‐frame video forgery detection strategy based on a hybrid optimization tuned deep‐convolutional neural network (deep‐CNN) classifier. The significance of the proposed method lies in developing the proposed raven‐finch optimization algorithm that tunes the weights of the deep‐CNN to exhibit enhanced detection accuracy. The proposed raven‐finch optimization algorithm combines raven search agents and finches search agents, possessing the benefits of both search agents. The features of the input video frames act as the input to the deep‐CNN classifier that detects forgery. The performance of the proposed raven‐finch‐based deep CNN method is analyzed in terms of the performance indices, such as accuracy, sensitivity, and specificity. It is attained to be 97.56%, 95.48%, and 96.38%, respectively, which shows the superiority of the proposed method for intra‐frame forgery detection.