An hybrid deep learning approach for depression prediction from user tweets using feature-rich CNN and bi-directional LSTM

Multimedia Tools and Applications, Journal Year: 2022, Volume and Issue: 81(17), P. 23649 - 23685

Published: March 18, 2022

Language: Английский

---

From Treatment to Healing:Envisioning a Decolonial Digital Mental Health

CHI Conference on Human Factors in Computing Systems, Journal Year: 2022, Volume and Issue: unknown, P. 1 - 23

Published: April 28, 2022

The field of digital mental health is making strides in the application technology to broaden access care. We critically examine how these technology-mediated forms care might amplify historical injustices, and erase minoritized experiences expressions distress illness. draw on decolonial thought critiques identity-based algorithmic bias analyze underlying power relations impacting technologies today, envision new pathways towards a health. argue that one centers lived experience over rigid classification, conscious structural factors influence wellbeing, fundamentally designed deter creation differentials prevent people from having agency their Stemming this vision, we make recommendations for researchers designers can support more equitable futures experiencing

Language: Английский

---

Depression detection using emotional artificial intelligence and machine learning: A closer review

Materials Today Proceedings, Journal Year: 2022, Volume and Issue: 58, P. 217 - 226

Published: Jan. 1, 2022

Language: Английский

---

Ensemble Multifeatured Deep Learning Models and Applications: A Survey

IEEE Access, Journal Year: 2023, Volume and Issue: 11, P. 107194 - 107217

Published: Jan. 1, 2023

Ensemble multifeatured deep learning methodology has emerged as a powerful approach to overcome the limitations of single models in terms generalization, robustness, and performance. This survey provides an extended review ensemble models, their applications, challenges, future directions. We explore potential applications these across various domains, including computer vision, medical imaging, natural language processing, speech recognition. By combining strengths multiple features, have demonstrated improved performance adaptability diverse problem settings. also discuss challenges associated with such model interpretability, computational complexity, selection, adversarial personalized federated learning. highlights recent advancements addressing emphasizes importance continued research tackling issues enable widespread adoption models. It outlook on directions, focusing development new algorithms, frameworks, hardware architectures that can efficiently handle large-scale computations required by Moreover, it underlines need for better understanding trade-offs between accuracy, resources optimize design deployment

Language: Английский

---

Exploring emotional patterns in social media through NLP models to unravel mental health insights

Healthcare Technology Letters, Journal Year: 2025, Volume and Issue: 12(1)

Published: Jan. 1, 2025

Abstract This study aimed to develop an advanced ensemble approach for automated classification of mental health disorders in social media posts. The research question was: can fine‐tuned transformer models (XLNet, RoBERTa, and ELECTRA) with Bayesian hyperparameter optimization improve the accuracy disorder text. Three were on a dataset posts labelled 15 distinct disorders. was employed tuning, optimizing learning rate, number epochs, gradient accumulation steps, weight decay. A voting then implemented combine predictions individual models. proposed achieved highest 0.780, outperforming models: XLNet (0.767), RoBERTa (0.775), ELECTRA (0.755). approach, integrating XLNet, optimization, demonstrated improved classifying from method shows promise enhancing digital potentially aiding early detection intervention strategies. Future work should focus expanding dataset, exploring additional techniques, investigating model's performance across different platforms languages.

Language: Английский

---

What Does Your Bio Say? Inferring Twitter Users’ Depression Status From Multimodal Profile Information Using Deep Learning

IEEE Transactions on Computational Social Systems, Journal Year: 2021, Volume and Issue: 9(5), P. 1484 - 1494

Published: Oct. 13, 2021

People suffering from stress and various mental health problems find it easier to express share their feelings on online platforms, such as Twitter. However, the imposed character limit (280 characters) by Twitter infrequent activities of a section users poses serious setback in using computational methods for analysis or emotion research. provides rich metadata information about its (such user's description, geolocation, profile image URL), which can provide valuable regarding state users. We hypothesize that Twitter's some depression cues, may help an early low-profile evaluation. In this article, we investigate hypothesis developing end-to-end multimodal multitask (MT) system detection (primary task) recognition (auxiliary task), where variation based different user descriptions assists learning primary task. The proposed attains 70% accuracy task outperforming several single-task (ST) baselines built combination input features. Our findings indicate Twitters's be leveraged detect among with significant confidence.

Language: Английский

---

An optimized deep learning approach for suicide detection through Arabic tweets

PeerJ Computer Science, Journal Year: 2022, Volume and Issue: 8, P. e1070 - e1070

Published: Aug. 23, 2022

Many people worldwide suffer from mental illnesses such as major depressive disorder (MDD), which affect their thoughts, behavior, and quality of life. Suicide is regarded the second leading cause death among teenagers when treatment not received. Twitter a platform for expressing emotions thoughts about many subjects. studies, including this one, suggest using social media data to track depression other illnesses. Even though Arabic widely spoken has complex syntax, detection methods have been applied language. The tweets dataset should be scraped annotated first. Then, complete framework categorizing tweet inputs into two classes (such Normal or Suicide) suggested in study. article also proposes an preprocessing algorithm that contrasts lemmatization, stemming, various lexical analysis methods. Experiments are conducted Internet. Five different annotators data. Performance metrics reported on latest Bidirectional Encoder Representations Transformers (BERT) Universal Sentence (USE) models. measured performance balanced accuracy, specificity, F1-score, IoU, ROC, Youden Index, NPV, weighted sum metric (WSM). Regarding USE models, best-weighted (WSM) 80.2%, with regards BERT best WSM 95.26%.

Language: Английский

---

A survey on detecting mental disorders with natural language processing: Literature review, trends and challenges

Computer Science Review, Journal Year: 2024, Volume and Issue: 53, P. 100654 - 100654

Published: June 22, 2024

Language: Английский

---

Early detection of depression on Twitter using machine learning techniques and pre-processing steps

AIP conference proceedings, Journal Year: 2025, Volume and Issue: 3224, P. 030003 - 030003

Published: Jan. 1, 2025

Language: Английский

---

MBTI Personality Profiling from Tweets Using Machine Learning

Lecture notes in electrical engineering, Journal Year: 2025, Volume and Issue: unknown, P. 309 - 323

Published: Jan. 1, 2025

Language: Английский

---