東華大學圖書館 |

Language: English

Help

回圖書館首頁

手機版館藏查詢

Back

Switch To: Labeled | MARC Mode | ISBD

Machine Learning Methods in Social M...

Ricard, Benjamin Joseph,

Linked to FindBook

Google Book

Amazon

博客來

Machine Learning Methods in Social Media Analysis with Applications in Mental Health /

Record Type:	Electronic resources : Monograph/item
Title/Author:	Machine Learning Methods in Social Media Analysis with Applications in Mental Health // Benjamin Joseph Ricard.
Author:	Ricard, Benjamin Joseph,
Description:	1 electronic resource (123 pages)
Notes:	Source: Dissertations Abstracts International, Volume: 83-03, Section: B.
Contained By:	Dissertations Abstracts International83-03B.
Subject:	Epidemiology. -
Online resource:	https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28717160
ISBN:	9798538124800

Machine Learning Methods in Social Media Analysis with Applications in Mental Health /
Ricard, Benjamin Joseph,

Machine Learning Methods in Social Media Analysis with Applications in Mental Health /Benjamin Joseph Ricard. - 1 electronic resource (123 pages)

Source: Dissertations Abstracts International, Volume: 83-03, Section: B.

Mental health disorders, including depression and alcohol abuse, are underdiagnosed and undertreated, directly contributing to increased mortality through suicide and heart disease. Traditional mental health diagnostic tests have been criticized for their lack of reliability and generalizability. Social media analysis provides a promising direction for mental health research, particularly for improving screening and intervention strategies. Data-driven machine learning models have demonstrated utility in processing social media information to identify mental health disorders. However, previous empirical analyses of texts and images have been limited by the lack of inclusion of information from comments, hashtags, and themed communities to examine individual and population-level mental health outcomes. In this dissertation, we proposed a textual analysis of community-generated data (i.e., comments) to predict depression in Instagram in users. Our results confirmed statistically significant signals that indicated information in comments written by a user's friends to detect depression in that user. Additionally, combining community and user-generated models outperformed physician benchmarks and either model alone. In a second study, we examined the utility of hashtags and themed communities on Reddit and Twitter to detect population-level alcohol abuse. A Bidirectional Encoder Representations from Transformers (BERT) trained on Reddit alcohol-related communities identified Twitter hashtags that significantly correlate to population-level alcohol-related outcomes in the United States. Finally, a Convolutional Neural Network (CNN) and BERT model was trained on images and texts from depression-related subreddit communities and evaluated using international and multi-lingual Twitter data. Analysis indicated that the predictions from the image and text model results had a significant correlation with depression-related outcomes at the country level. These and similar methods can improve the analysis of mental health outcomes using social media data.

English

ISBN: 9798538124800Subjects--Topical Terms:

568544
Epidemiology.
Subjects--Index Terms:

Instagram

Machine Learning Methods in Social Media Analysis with Applications in Mental Health /
LDR:03618nmm a22004813i 4500 001 2431388
005 20250616055137.5
006 m o d
007 cr|nu||||||||
008 261011s2021 miu||||||m |||||||eng d
020 $a 9798538124800
035 $a (MiAaPQD)AAI28717160
035 $a AAI28717160
040 $a MiAaPQD $b eng $c MiAaPQD $e rda
100 1 $a Ricard, Benjamin Joseph, $e author. $3 3820167
245 1 0 $a Machine Learning Methods in Social Media Analysis with Applications in Mental Health / $c Benjamin Joseph Ricard.
264 1 $a Ann Arbor : $b ProQuest Dissertations & Theses, $c 2021
300 $a 1 electronic resource (123 pages)
336 $a text $b txt $2 rdacontent
337 $a computer $b c $2 rdamedia
338 $a online resource $b cr $2 rdacarrier
500 $a Source: Dissertations Abstracts International, Volume: 83-03, Section: B.
500 $a Advisors: Hassanpour, Saeed Committee members: Marsch, Lisa; Danforth, Chris; MacKenzie, Todd.
502 $b Ph.D. $c Dartmouth College $d 2021.
520 $a Mental health disorders, including depression and alcohol abuse, are underdiagnosed and undertreated, directly contributing to increased mortality through suicide and heart disease. Traditional mental health diagnostic tests have been criticized for their lack of reliability and generalizability. Social media analysis provides a promising direction for mental health research, particularly for improving screening and intervention strategies. Data-driven machine learning models have demonstrated utility in processing social media information to identify mental health disorders. However, previous empirical analyses of texts and images have been limited by the lack of inclusion of information from comments, hashtags, and themed communities to examine individual and population-level mental health outcomes. In this dissertation, we proposed a textual analysis of community-generated data (i.e., comments) to predict depression in Instagram in users. Our results confirmed statistically significant signals that indicated information in comments written by a user's friends to detect depression in that user. Additionally, combining community and user-generated models outperformed physician benchmarks and either model alone. In a second study, we examined the utility of hashtags and themed communities on Reddit and Twitter to detect population-level alcohol abuse. A Bidirectional Encoder Representations from Transformers (BERT) trained on Reddit alcohol-related communities identified Twitter hashtags that significantly correlate to population-level alcohol-related outcomes in the United States. Finally, a Convolutional Neural Network (CNN) and BERT model was trained on images and texts from depression-related subreddit communities and evaluated using international and multi-lingual Twitter data. Analysis indicated that the predictions from the image and text model results had a significant correlation with depression-related outcomes at the country level. These and similar methods can improve the analysis of mental health outcomes using social media data.
546 $a English
590 $a School code: 0059
650 4 $a Epidemiology. $3 568544
650 4 $a Mental depression. $3 3556007
650 4 $a Addictive behaviors. $3 3563047
650 4 $a Mental disorders. $3 728624
650 4 $a Medical research. $2 bicssc $3 1556686
650 4 $a Artificial intelligence. $3 516317
650 4 $a Public health. $3 534748
650 4 $a Web studies. $3 2122754
650 4 $a Mental health. $3 534751
650 4 $a Websites. $2 gtt $3 1062467
653 $a Instagram
653 $a Machine learning
653 $a Reddit
653 $a Social media
653 $a Twitter
653 $a Alcohol abuse
653 $a Depression screening
690 $a 0766
690 $a 0573
690 $a 0800
690 $a 0646
690 $a 0347
710 2 $a Dartmouth College. $b Quantitative Biomedical Sciences. $e degree granting institution. $3 3820168
720 1 $a Hassanpour, Saeed $e degree supervisor.
773 0 $t Dissertations Abstracts International $g 83-03B.
790 $a 0059
791 $a Ph.D.
792 $a 2021
856 4 0 $u https://pqdd.sinica.edu.tw/twdaoapp/servlet/advanced?query=28717160