Identifying the Risk Factors of Allergic Rhinitis Based on Zhihu Comment Data Using a Topic-Enhanced Word-Embedding Model: Mixed Method Study and Cluster Analysis (Preprint) DOI
Dongxiao Gu, Q. Z. Wang, Yidong Chai

et al.

Published: April 19, 2023

BACKGROUND Allergic rhinitis (AR) is a chronic disease, and several risk factors predispose individuals to the condition in their daily lives, including exposure allergens inhalation irritants. Analyzing potential that can trigger AR provide reference material for use reduce its occurrence lives. Nowadays, social media part of life, with an increasing number people using at least 1 platform regularly. Social enables users share experiences among large groups who same interests experience afflictions. Notably, these channels promote ability health information. OBJECTIVE This study aims construct intelligent method (TopicS-ClusterREV) identifying based on comments. The main questions were as follows: How many comments contained factor information? categories be summarized into? do AR? METHODS crawled all data from May 2012 2022 under topic <i>allergic rhinitis</i> Zhihu, obtaining total 9628 posts 33,747 We improved Skip-gram model train topic-enhanced word vector representations (TopicS) then vectorized annotated text items training classifier. Furthermore, cluster analysis enabled closer look into opinions expressed category, namely gaining insight how AR. RESULTS Our classifier identified more containing than other classification models, accuracy rate 96.1% recall 96.3%. In general, we clustered texts 28 categories, season, region, mites being most common factors. gained each category; example, seasonal changes increased temperature differences between day night disrupt body’s immune system lead development allergies. CONCLUSIONS approach handle amount extract effectively. Moreover, summary serve experimental also pathway triggers finding guide management plans interventions

Language: Английский

Intelligent evaluation system for new energy vehicles based on sentiment analysis: An MG-PL-3WD method DOI
Chao Zhang,

Qifei Wen,

Deyu Li

et al.

Engineering Applications of Artificial Intelligence, Journal Year: 2024, Volume and Issue: 133, P. 108485 - 108485

Published: April 25, 2024

Language: Английский

Citations

9
Identifying the Risk Factors of Allergic Rhinitis Based on Zhihu Comment Data Using a Topic-Enhanced Word-Embedding Model: Mixed Method Study and Cluster Analysis DOI Creative Commons
Dongxiao Gu, Q. Z. Wang, Yidong Chai

et al.

Journal of Medical Internet Research, Journal Year: 2024, Volume and Issue: 26, P. e48324 - e48324

Published: Feb. 22, 2024

Background Allergic rhinitis (AR) is a chronic disease, and several risk factors predispose individuals to the condition in their daily lives, including exposure allergens inhalation irritants. Analyzing potential that can trigger AR provide reference material for use reduce its occurrence lives. Nowadays, social media part of life, with an increasing number people using at least 1 platform regularly. Social enables users share experiences among large groups who same interests experience afflictions. Notably, these channels promote ability health information. Objective This study aims construct intelligent method (TopicS-ClusterREV) identifying based on comments. The main questions were as follows: How many comments contained factor information? categories be summarized into? do AR? Methods crawled all data from May 2012 2022 under topic allergic Zhihu, obtaining total 9628 posts 33,747 We improved Skip-gram model train topic-enhanced word vector representations (TopicS) then vectorized annotated text items training classifier. Furthermore, cluster analysis enabled closer look into opinions expressed category, namely gaining insight how AR. Results Our classifier identified more containing than other classification models, accuracy rate 96.1% recall 96.3%. In general, we clustered texts 28 categories, season, region, mites being most common factors. gained each category; example, seasonal changes increased temperature differences between day night disrupt body’s immune system lead development allergies. Conclusions approach handle amount extract effectively. Moreover, summary serve experimental also pathway triggers finding guide management plans interventions

Language: Английский

Citations

6
Getting better? Examining the effects of social support in OHCs on users’ emotional improvement DOI
Yuehua Zhao, Linyi Zhang

Information Processing & Management, Journal Year: 2024, Volume and Issue: 61(4), P. 103754 - 103754

Published: April 26, 2024

Language: Английский

Citations

3
Analyzing the impact of unemployment on mental health among Chinese university graduates: a study of emotional and linguistic patterns on Weibo DOI Creative Commons
Miaoqing Tan, Zhigang Wu, Li Jin

et al.

Frontiers in Public Health, Journal Year: 2024, Volume and Issue: 12

Published: May 9, 2024

Purpose This study explores the intricate relationship between unemployment rates and emotional responses among Chinese university graduates, analyzing how these factors correlate with specific linguistic features on popular social media platform Sina Weibo. The goal is to uncover patterns that elucidate psychological dimensions of challenges this demographic. Methods analysis utilized a dataset 30,540 Weibo posts containing keywords related anxiety, collected from January 2019 June 2023. were pre-processed eliminate noise refine data quality. Linear regression textual analyses employed identify correlations for individuals aged 16–24 characteristics posts. Results found significant fluctuations in urban youth rates, peaking at 21.3% A corresponding increase anxiety-related expressions was noted posts, peak aligning high rates. Linguistic revealed category “Affect” showed strong positive correlation indicating increased expression alongside rising unemployment. Other categories such as “Negative emotion” “Sadness” also correlations, highlighting robust economic distress. Conclusion findings underscore profound impact well-being students, suggesting hardships are closely linked stress heightened negative emotions. contributes holistic understanding socio-economic faced by young adults, advocating comprehensive support systems address both facets

Language: Английский

Citations

3
Machine Learning for Depression Risk Monitoring on Chinese Social Media: A Comprehensive Evaluation and Analysis (Preprint) DOI Creative Commons
Zhenwen Zhang,

Zepeng Li,

Zhihua Guo

et al.

Published: March 11, 2024

BACKGROUND Depression is a significant global public health issue that affects the physical and mental well-being of hundreds millions people worldwide. However, substantial number individuals with depression on social media often go undiagnosed struggle to access timely effective treatment, increasingly becoming major societal concern. OBJECTIVE This paper aims explore develop an online risk detection method based deep learning technology identify at Chinese platform Sina Weibo. METHODS We initially collected approximately 527,333 posts publicly shared over one year from 1600 without Weibo platform. Subsequently, we developed hierarchical Transformer network learn semantic features for each user. comprises two levels structures, word level other sentence level. These Transformers are employed extract textual post, aggregated all user generate user-level features. A classifier then applied predict depression. Finally, conducted statistical linguistic analyses content using LIWC. RESULTS divided original dataset into training, validation, test sets. The training set consists 1000 100 validation includes 600 users, 300 Our achieved accuracy 84.62%, precision 84.43%, recall 84.50%, F1 score 84.32% applying sampling techniques. After our proposed retrieval-based strategy, 95.46%, 95.30%, 95.70%, 95.43%. results strongly demonstrate effectiveness superiority model technique. provides new insights large-scale through media. Through language behavior analysis, it observed more likely use negation words (the value "swear" 0.001253). may indicate presence negative emotions, rejection, doubt, disagreement, or aversion expressed by Additionally, also found tend emotional vocabulary in their expressions (NegEmo: 0.022306, Anx: 0.003829, Anger: 0.004327, Sad: 0.005740), which reflect internal emotions psychological state. frequent could be way express feelings towards life, themselves, surrounding environment. CONCLUSIONS research feasibility methods detecting potential large-scale, automated, non-invasive prediction among users

Language: Английский

Citations

0
Natural Language Processing for Depression Prediction on Sina Weibo: Method Study and Analysis DOI Creative Commons
Zhenwen Zhang,

Jianghong Zhu,

Zhihua Guo

et al.

JMIR Mental Health, Journal Year: 2024, Volume and Issue: 11, P. e58259 - e58259

Published: June 4, 2024

Abstract Background Depression represents a pressing global public health concern, impacting the physical and mental well-being of hundreds millions worldwide. Notwithstanding advances in clinical practice, an alarming number individuals at risk for depression continue to face significant barriers timely diagnosis effective treatment, thereby exacerbating burgeoning social crisis. Objective This study seeks develop novel online detection method using natural language processing technology identify on Chinese media platform Sina Weibo. Methods First, we collected approximately 527,333 posts publicly shared over 1 year from 1600 with without Weibo platform. We then developed hierarchical transformer network learning user-level semantic representations, which consists 3 primary components: word-level encoder, post-level aggregation encoder. The encoder learns embeddings individual posts, while explores features user post sequences. aggregates sequence semantics generate representation that can be classified as depressed or nondepressed. Next, classifier is employed predict depression. Finally, conducted statistical linguistic analyses content Linguistic Inquiry Word Count. Results divided original data set into training, validation, test sets. training consisted 1000 Similarly, each validation comprised 600 users, 300 both cohorts (depression nondepression). Our achieved accuracy 84.62%, precision 84.43%, recall 84.50%, F -score 84.32% employing sampling techniques. However, by applying our proposed retrieval-based strategy, observed improvements performance: 95.46%, 95.30%, 95.70%, 95.43%. These outstanding results clearly demonstrate effectiveness superiority model technique. breakthrough provides new insights large-scale through media. Through behavior analysis, discovered are more likely use negation words (the value “swear” 0.001253). may indicate presence negative emotions, rejection, doubt, disagreement, aversion Additionally, analysis revealed tend emotional vocabulary their expressions (“NegEmo”: 0.022306; “Anx”: 0.003829; “Anger”: 0.004327; “Sad”: 0.005740), reflect internal emotions psychological state. frequent could way express feelings toward life, themselves, surrounding environment. Conclusions research feasibility deep methods detect findings provide potential large-scale, automated, noninvasive prediction among users.

Language: Английский

Citations

0
Predicting negative attitudes towards suicide in social media texts: prediction model development and validation study DOI Creative Commons
Ang Li

Frontiers in Public Health, Journal Year: 2024, Volume and Issue: 12

Published: July 8, 2024

Background Implementing machine learning prediction of negative attitudes towards suicide may improve health outcomes. However, in previous studies, varied forms were not adequately considered, and developed models lacked rigorous external validation. By analyzing a large-scale social media dataset (Sina Weibo), this paper aims to fully cover develop classification model for predicting as whole, then externally validate its performance on population individual levels. Methods 938,866 Weibo posts with relevant keywords downloaded, including 737,849 updated between 2009 2014 ( 2009–2014 ), 201,017 2015 2020 2015–2020 ). (1) For development, based 10,000 randomly selected from , human-based content analysis was performed manually determine labels each post (non-negative or attitudes). Then, computer-based conducted automatically extract psycholinguistic features the same posts. Finally, features. (2) validation, level, implemented remaining 727,849 validated by comparing proportions predicted human-coded results. Besides, similar analyses 300 actual Results F1 area under ROC curve (AUC) values reached 0.93 0.97. significant differences but very small effect sizes observed whole sample χ 2 1 = 32.35, p &lt; 0.001; Cramer’s V 0.007, 0.001), men 9.48, 0.002; 0.005, 0.002), women 25.34, 0.009, 0.001). AUC 0.76 0.74. Conclusion This study demonstrates efficiency necessity confirms that validation is essential before implementing into practice.

Language: Английский

Citations

0
Sentiment Classification of Anxiety-Related Texts in Social Media via Fuzing Linguistic and Semantic Features DOI

Jianghong Zhu,

Zhenwen Zhang, Zhihua Guo

et al.

IEEE Transactions on Computational Social Systems, Journal Year: 2024, Volume and Issue: 11(5), P. 6819 - 6829

Published: July 18, 2024

Language: Английский

Citations

0
Changes in stigma and population mental health literacy before and after the Covid-19 pandemic: analyses of repeated cross-sectional studies DOI Creative Commons
Petr Winkler,

Benjamin Kunc,

Zoe Guerrero

et al.

SSM - Mental Health, Journal Year: 2024, Volume and Issue: unknown, P. 100369 - 100369

Published: Nov. 1, 2024

Language: Английский

Citations

0
Anxiety management strategies on TikTok: a discourse analysis of collective coping mechanisms DOI
Yavuz Selim Balcıoğlu

Journal of Mental Health, Journal Year: 2024, Volume and Issue: 33(5), P. 619 - 629

Published: Aug. 19, 2024

This study investigates the discourse on anxiety management strategies within TikTok platform, analyzing a substantial dataset of 45,639 comments collected over year.

Language: Английский

Citations

0