Graph contrastive learning (GCL) is a promising direction toward alleviating the label dependence, poor generalization and weak robustness of graph neural networks, learning representations with invariance, and discriminability by solving pretasks. The pretasks are mainly built on mutual information estimation, which requires data augmentation to construct positive samples with similar semantics to learn invariant signals and negative samples with dissimilar semantics to empower representation discriminability. However, an appropriate data augmentation configuration depends heavily on lots of empirical trials such as choosing the compositions of data augmentation techniques and the corresponding hyperparameter settings. We propose an augmentation-free GCL method, invariant-discriminative GCL (iGCL), that does not intrinsically require negative samples. iGCL designs the invariant-discriminative loss (ID loss) to learn invariant and discriminative representations. On the one hand, ID loss learns invariant signals by directly minimizing the mean square error (MSE) between the target samples and positive samples in the representation space. On the other hand, ID loss ensures that the representations are discriminative by an orthonormal constraint forcing the different dimensions of representations to be independent of each other. This prevents representations from collapsing to a point or subspace. Our theoretical analysis explains the effectiveness of ID loss from the perspectives of the redundancy reduction criterion, canonical correlation analysis (CCA), and information bottleneck (IB) principle. The experimental results demonstrate that iGCL outperforms all baselines on five node classification benchmark datasets. iGCL also shows superior performance for different label ratios and is capable of resisting graph attacks, which indicates that iGCL has excellent generalization and robustness. The source code is available at https://github.com/lehaifeng/ T-GCN/tree/master/iGCL.

Feelings of inferiority are complex emotions that usually indicate perceived weakness and helplessness. A lack of timely and effective interventions may bring serious consequences to individuals with inferiority feelings. Due to privacy concerns, those people often hesitate to seek face-to-face help, but they usually spontaneously share their feelings on social media, which makes social media a good resource for ample inferiority-related data. We randomly selected a sample of posts indicating inferiority feelings to explore the causes of inferiority. Through language analysis and natural language processing, we constructed a Word2Vec model of inferiority based on social media data and applied it to the cause analysis of inferiority feelings. The main causes of inferiority feelings are personal experience, social interaction, love relationship, etc. People feeling inferior about their personal experiences usually are largely influenced by their ways of thinking and life attitudes. Social and emotional factors overlap somewhat in the development of inferiority. In love relationships, males are more prone to inferiority feeling than females. These findings will help relevant institutions and organizations better understand people with inferiority feelings and facilitate the development of targeted treatment for those with potential self-esteem problems.

Insomnia is a prevalent sleep disorder that causes serious harm to individuals and society. It is closely linked to not only personal factors but also social, economic and other factors. This study explores the influencing factors and spatial differentiation of insomnia from the perspective of social media. This paper chose China's largest social media platform, Sina Weibo, as its data source. Then, based on the collected relevant data of 288 Chinese cities from 2013 to 2017, it explored the impact of economic, social, and environmental factors and an educated population on insomnia. Additionally, the importance and interaction of each influencing factor were analyzed. According to the results, the gross domestic product (GDP), proportion of households connected to the Internet and number of students in regular institutions of higher education are the major factors that influence insomnia, and their influences show obvious spatial nonstationarity. Rapid GDP growth has increased the probability of insomnia, and the positive correlation between the proportion of households connected to the internet and insomnia has strengthened annually. Although the impact of insomnia on college students decreased in some regions, the overall impact was still increasing annually, and spatial nonstationarity was obvious. Properly controlling GDP growth and unnecessary time spent online and guiding people to develop healthy Internet surfing habits and lifestyles will help improve their sleep quality. Our research results will help relevant professionals better understand the distribution of regional insomnia and provide a reference for related departments to formulate regional insomnia prevention and treatment policies.

Colorectal cancer is a major public health problem for its contribution to morbidity and mortality globally, and one of its major inducements is diet. Despite the achievements in the correlation studies about dietary habit and colorectal cancer, as an important component of dietary habit, taste preference has not received enough attention, its quantitative relationship with colorectal cancer has not been fully revealed. Using data from crowdsourced recipes, restaurant points of interest, and regional colorectal cancer mortality rates, this study quantitatively analysed the effects of different tastes on colorectal cancer, established a spatio-temporal link between taste and the disease, elucidating taste’s role in the spatial distribution of colorectal cancer. According to the analysis, Chinese taste preference presents significant spatial heterogeneity, and Chinese cuisines most favour ’Spicy’ which makes up the top two risk factors of colorectal cancer with ’Pungent’. Also, the interactions of Salty’ with other tastes might increase colorectal cancer risk, the same goes for ’Oily’. The proposed quantitative analysis method based on crowdsourcing data can be applied to researches on other diseases, and the study can provide a new method and a new perspective for relevant studies.