scholarly journals Cross-Domain Depression Detection via Harvesting Social Media

2018 ◽  
Author(s):  
Tiancheng Shen ◽  
Jia Jia ◽  
Guangyao Shen ◽  
Fuli Feng ◽  
Xiangnan He ◽  
...  
Keyword(s):  
Social Media ◽  
Poor Performance ◽  
Well Being ◽  
Relevant Information ◽  
Target Domain ◽  
Combination Strategy ◽  
Cross Domain ◽  
Depression Detection ◽  
Improved Performance ◽  
Model Training

Depression detection is a significant issue for human well-being. In previous studies, online detection has proven effective in Twitter, enabling proactive care for depressed users. Owing to cultural differences, replicating the method to other social media platforms, such as Chinese Weibo, however, might lead to poor performance because of insufficient available labeled (self-reported depression) data for model training. In this paper, we study an interesting but challenging problem of enhancing detection in a certain target domain (e.g. Weibo) with ample Twitter data as the source domain. We first systematically analyze the depression-related feature patterns across domains and summarize two major detection challenges, namely isomerism and divergency. We further propose a cross-domain Deep Neural Network model with Feature Adaptive Transformation & Combination strategy (DNN-FATC) that transfers the relevant information across heterogeneous domains. Experiments demonstrate improved performance compared to existing heterogeneous transfer methods or training directly in the target domain (over 3.4% improvement in F1), indicating the potential of our model to enable depression detection via social media for more countries with different cultural settings.

Fusion of part-of-speech vectors and attention mechanisms for cross-domain sentiment analysis

2021 ◽  
pp. 1-9
Author(s):  
Ting Lu ◽  
Yan Xiang ◽  
Junge Liang ◽  
Li Zhang ◽  
Mingfang Zhang
Keyword(s):  
Sentiment Analysis ◽  
Poor Performance ◽  
Feature Representation ◽  
Grand Challenge ◽  
Feature Mapping ◽  
Target Domain ◽  
Specific Domain ◽  
Part Of Speech ◽  
Cross Domain ◽  
Speech Information

The grand challenge of cross-domain sentiment analysis is that classifiers trained in a specific domain are very sensitive to the discrepancy between domains. A sentiment classifier trained in the source domain usually have a poor performance in the target domain. One of the main strategies to solve this problem is the pivot-based strategy, which regards the feature representation as an important component. However, part-of-speech information was not considered to guide the learning of feature representation and feature mapping in previous pivot-based models. Therefore, we present a fused part-of-speech vectors and attention-based model (FAM). In our model, we fuse part-of-speech vectors and feature word embeddings as the representation of features, giving deep semantics to mapping features. And we adopt Multi-Head attention mechanism to train the cross-domain sentiment classifier to obtain the connection between different features. The results of 12 groups comparative experiments on the Amazon dataset demonstrate that our model outperforms all baseline models in this paper.

scholarly journals Deep Transfer Learning for Biology Cross-Domain Image Classification

2021 ◽  
Vol 2021 ◽  
pp. 1-19
Author(s):  
Chunfeng Guo ◽  
Bin Wei ◽  
Kun Yu
Keyword(s):  
Image Classification ◽  
Transfer Learning ◽  
Poor Performance ◽  
Target Domain ◽  
Deep Convolutional Neural Networks ◽  
Specific Domain ◽  
Source Domain ◽  
Cross Domain ◽  
Domain Transfer ◽  
Domain Transfer Learning

Automatic biology image classification is essential for biodiversity conservation and ecological study. Recently, due to the record-shattering performance, deep convolutional neural networks (DCNNs) have been used more often in biology image classification. However, training DCNNs requires a large amount of labeled data, which may be difficult to collect for some organisms. This study was carried out to exploit cross-domain transfer learning for DCNNs with limited data. According to the literature, previous studies mainly focus on transferring from ImageNet to a specific domain or transferring between two closely related domains. While this study explores deep transfer learning between species from different domains and analyzes the situation when there is a huge difference between the source domain and the target domain. Inspired by the analysis of previous studies, the effect of biology cross-domain image classification in transfer learning is proposed. In this work, the multiple transfer learning scheme is designed to exploit deep transfer learning on several biology image datasets from different domains. There may be a huge difference between the source domain and the target domain, causing poor performance on transfer learning. To address this problem, multistage transfer learning is proposed by introducing an intermediate domain. The experimental results show the effectiveness of cross-domain transfer learning and the importance of data amount and validate the potential of multistage transfer learning.

Examining the accuracy of estimated smartphone use: How well-being and usage level predict discrepancies between estimated and actual use

2019 ◽  
Author(s):  
Craig Sewall ◽  
Daniel Rosen ◽  
Todd M. Bear
Keyword(s):  
Public Health ◽  
Social Media ◽  
Life Satisfaction ◽  
Mobile Device ◽  
Screen Time ◽  
Well Being ◽  
Actual Use ◽  
Smartphone Use ◽  
Usage Level

The increasing ubiquity of mobile device and social media (SM) use has generated a substantial amount of research examining how these phenomena may impact public health. Prior studies have found that mobile device and SM use are associated with various aspects of well-being. However, a large portion of these studies relied upon self-reported estimates to measure amount of use, which can be inaccurate. Utilizing Apple’s “Screen Time” application to obtain actual iPhone and SM use data, the current study examined the accuracy of self-reported estimates, how inaccuracies bias relationships between use and well-being (depression, loneliness, and life satisfaction), and the degree to which inaccuracies were predicted by levels of well-being. Among a sample of 393 iPhone users, we found that: a.) participants misestimated their weekly overall iPhone and SM use by 22.1 and 16.6 hours, respectively; b.) the correlations between estimated use and well-being variables were consistently stronger than the correlations between actual use and well-being variables; and c.) the amount of inaccuracy in estimated use is associated with levels of participant well-being as well as amount of use. These findings suggest that estimates of device/SM use may be biased by factors that are fundamental to the relationships being investigated. **This manuscript is currently under review**

Measuring Relevant Information in Health Social Network Conversations and Clinical Diagnosis Cases (Preprint)

2018 ◽  
Author(s):  
Albert Moreira ◽  
Raul Alonso-Calvo ◽  
Alberto Muñoz ◽  
Jose Crespo
Keyword(s):  
Social Media ◽  
Social Network ◽  
Relevant Information ◽  
Online Discussions ◽  
Biomedical Knowledge ◽  
Expert Opinions ◽  
Health Related ◽  
Health Social Network ◽  
Discussion System ◽  
Clinical Cases

BACKGROUND Internet and Social media is an enormous source of information. Health Social Networks and online collaborative environments enable users to create shared content that afterwards can be discussed. While social media discussions for health related matters constitute a potential source of knowledge, characterizing the relevance of participations from different users is a challenging task. OBJECTIVE The aim of this paper is to present a methodology designed for quantifying relevant information provided by different participants in clinical online discussions. METHODS A set of key indicators for different aspects of clinical conversations and specific clinical contributions within a discussion have been defined. These indicators make use of biomedical knowledge extraction based on standard terminologies and ontologies. These indicators allow measuring the relevance of information of each participant of the clinical conversation. RESULTS Proposed indicators have been applied to two discussions extracted from PatientsLikeMe, as well as to two real clinical cases from the Sanar collaborative discussion system. Results obtained from indicators in the tested cases have been compared with clinical expert opinions to check indicators validity. CONCLUSIONS The methodology has been successfully used for describing participant interactions in real clinical cases belonging to a collaborative clinical case discussion tool and from a conversation from a Health Social Network.

Revolutionizing Mental Health with Social Media

2019 ◽  
pp. 259-284
Author(s):  
Munmun De Choudhury
Keyword(s):  
Social Media ◽  
Well Being ◽  
Linguistic Expression ◽  
Ethical Considerations ◽  
Self Disclosure ◽  
Vulnerable Communities ◽  
Tailored Interventions ◽  
Challenges And Opportunities ◽  
Social Media Platforms

Social media platforms have emerged as rich repositories of information relating to people’s activities, emotions, and linguistic expression. This chapter highlights how these data may be harnessed to reason about human mental and psychological well-being. It also discusses the emergent role of social media in providing a platform of self-disclosure and support to distressed and vulnerable communities. It reflects on how this new line of research bears potential for informing the design of timely and tailored interventions, provisions for improved personal and societal well-being assessment, privacy and ethical considerations, and the challenges and opportunities of the increasing ubiquity of social media.

scholarly journals Young adults’ coping strategies against loneliness during the COVID-19-related quarantine in Greece

2021 ◽  
Author(s):  
Adrianos Golemis ◽  
Panteleimon Voitsidis ◽  
Eleni Parlapani ◽  
Vasiliki A Nikopoulou ◽  
Virginia Tsipropoulou ◽  
...  
Keyword(s):  
Social Media ◽  
Young Adults ◽  
Social Responsibility ◽  
Coping Strategies ◽  
Online Survey ◽  
Adult Population ◽  
Psychological Impact ◽  
Well Being ◽  
Daily Routine ◽  
Cross Sectional

Summary COVID-19 and the related quarantine disrupted young adults’ academic and professional life, daily routine and socio-emotional well-being. This cross-sectional study focused on the emotional and behavioural responses of a young adult population during the COVID-19-related quarantine in April 2020, in Greece. The study was conducted through an online survey. A total of 1559 young adults, aged 18−30 years, completed Steele’s Social Responsibility Motivation Scale and the De Jong Gierveld Loneliness Scale, and answered questions about compliance with instructions, quarantine-related behaviours and coping strategies. According to the results, participants displayed a relatively high sense of social responsibility (M = 16.09, SD = 2.13) and a trend towards moderate feeling of loneliness (M = 2.65, SD = 1.62); young women reported significantly higher levels of loneliness than men. The majority complied with instructions often (46.4%) or always (44.8%). Significantly more women created a new social media account and used the social media longer than 5 h/day, compared with men. Resorting to religion, practicing sports and sharing thoughts and feelings about COVID-19 with others predicted higher levels of social responsibility; humour, practicing sports and sharing thoughts and feelings about COVID-19 with others predicted lower levels of loneliness. Conclusively, COVID-19 is expected to have a significant psychological impact on young adults. Currently, Greece is going through the second quarantine period. This study raises awareness about loneliness in young adults during the COVID-19-related quarantine and highlights the importance of developing online programmes, attractive to younger people, to nurture adaptive coping strategies against loneliness.

scholarly journals Stand against anti-Asian racial discrimination during COVID-19: A call for action

2020 ◽  
pp. 002087282097061
Author(s):  
Qin Gao ◽  
Xiaofang Liu
Keyword(s):  
United States ◽  
Social Media ◽  
Racial Discrimination ◽  
Asian Americans ◽  
Well Being ◽  
Political Rhetoric ◽  
The United States ◽  
Psychological Well Being ◽  
Negative Impacts

Racial discrimination against people of Chinese and other Asian ethnicities has risen sharply in number and severity globally amid the COVID-19 pandemic. This rise has been especially rapid and severe in the United States, fueled by xenophobic political rhetoric and racist language on social media. It has endangered the lives of many Asian Americans and is likely to have long-term negative impacts on the economic, social, physical, and psychological well-being of Asian Americans. This essay reviews the prevalence and consequences of anti-Asian racial discrimination during COVID-19 and calls for actions in practice, policy, and research to stand against it.

Social Media and Emotional Well-being: Pursuit of Happiness or Pleasure

2021 ◽  
pp. 1326365X2110037
Author(s):  
D. Guna Graciyal ◽  
Deepa Viswam
Keyword(s):  
Social Media ◽  
Social Networking ◽  
Virtual World ◽  
Mixed Methods Research ◽  
Well Being ◽  
Pursuit Of Happiness ◽  
Social Dysfunction ◽  
Combining Data ◽  
Social Media Platforms ◽  
Depth Interviews

Virtual engagement of lives has been made possible with the advent of social media. Almost 80% of the day are spent virtually on Facebook, Instagram, Twitter, YouTube, Snapchat, etc. Usage of social media to connect to and communicate with the ones we care about is always healthy, termed as social networking. Social dysfunction occurs when the constant communication leads to the point where our real or offline life gets replaced by virtual or online life. There is a slight boundary between social networking and social dysfunction. When social networking is advantageous, social dysfunction affects emotional well-being. When emotional well-being is affected, many users experience a compulsion to dissociate from the real world as they find virtual world, full of fantasy and enjoyment. When the Internet was created, perhaps no one was aware of its potential. More than the convenience for sharing of information it has brought the world so close to crumbling the geographical boundaries. The more people-to-people communication is, the more is the strengthening of relationships, bonds grow stronger with ‘more’ social media platforms. Being on ‘more’ social media platforms has become a benchmark for living amidst the younger generation. Either as an activity of happiness or as an activity of pleasure, users tend to use social media at varying levels. This paper aims to conceptualize the the intricacies of social media in young lives and to discern whether their association is happiness or pleasure activity. The research method of this paper has a mixed-methods research design combining data from structured survey with information outputs from in-depth interviews.

Sign in / Sign up

Export Citation Format

Share Document