Behavior Variations and Their Implications for Popularity Promotions: From Elites to Mass on Weibo.

The boom in social media with regard to producing and consuming information simultaneously implies the crucial role of online user influence in determining content popularity. In particular, understanding behavior variations between the influential elites and the mass grassroots is an important issue in communication. However, how their behavior varies across user categories and content domains and how these differences influence content popularity are rarely addressed. From a novel view of seven content domains, a detailed picture of the behavior variations among five user groups, from the views of both the elites and mass, is drawn on Weibo, one of the most popular Twitter-like services in China. Interestingly, elites post more diverse content with video links, while the mass possess retweeters of higher loyalty. According to these variations, user-oriented actions for enhancing content popularity are discussed and testified. The most surprising finding is that the diverse content does not always bring more retweets, and the mass and elites should promote content popularity by increasing their retweeter counts and loyalty, respectively. For the first time, our results demonstrate the possibility of highly individualized strategies of popularity promotions in social media, instead of a universal principle.

Trading Imbalance in Chinese Stock Market-A High-Frequency View.

Although an imbalance of buying and selling profoundly affects the formation of market trends, a fine-granularity investigation of this perplexity of trading behavior is still missing. Instead of using existing entropy measures, this paper proposed a new indicator based on transaction dataset that enables us to inspect both the direction and the magnitude of this imbalance at high frequency, which we call "polarity". The polarity aims to measure the unevenness of the very essence trading desire based on the most micro decision making units. We investigate the relationship between the polarity and the return at both market-level and stock-level and find that the autocorrelated polarities cause a positive relation between lagged polarities and returns, while the current polarity is the opposite. It is also revealed that these associations shift according to the market conditions. In fact, when aggregating the one-minute polarities into daily signals, we find not only significant correlations disclosed by the market polarity and market emotion, but also the reliability of these signals in terms of reflecting the transitions of market-level behavior. These results imply that our presented polarity can reflect the market sentiment and condition in real time. Indeed, the trading polarity provides a new indicator from a high-frequency perspective to understand and foresee the market's behavior in a data-driven manner.

The long-term impacts of air quality on fine-grained online emotional responses to haze pollution in 160 Chinese cities.

Air pollution poses a great threat to public health and social stability by influencing multiple emotions. In particular, the air quality in developing countries is deteriorating along with rapid industrialization and urbanization, and multiple emotions may change along with regulation updates and air quality trending. Monitoring changes in public emotion is crucial for environmental governance. However, limited evidence exists for long-term effects of air quality on fine-grained emotions. Traditional surveys have the drawbacks of spatial limitations and high costs of time and money. Here, we use deep learning models to identify multiple emotions of over 10 million haze-related tweets and evaluate the effect of air quality on emotional predispositions for 160 cities from 2014 to 2019 in China. We find that sadness and joy are persistently associated with air quality, while anger and disgust are not. Surprisingly, the effects on fear vanished in the last three years. Moreover, air pollution initially had a greater impact on expressed fear in cities with higher income, poorer air quality and a greater percentage of women. Through popularity ranking and dynamic topic model, we interpretively revealed that people are no longer overly panicked and their attention is shifting toward policies and sources of haze. Our findings highlight the temporal evolution in the public's emotional response and provide significant implications for equitable public policies.

Weibo sentiments and stock return: A time-frequency view.

This study provides new insights into the relationships between social media sentiments and the stock market in China. Based on machine learning, we classify microblogs posted on Sina Weibo, a Twitter's variant in China into five detailed sentiments of anger, disgust, fear, joy, and sadness. Using wavelet analysis, we find close positive linkages between sentiments and the stock return, which have both frequency and time-varying features. Five detailed sentiments are positively related to the stock return for certain periods, particularly since October 2014 at medium to high frequencies of less than ten trading days, when the stock return is undergoing significant fluctuations. Sadness appears to have a closer relationship with the stock return than the other four sentiments. As to the lead-lag relationships, the stock return causes Weibo sentiments rather than reverse for most of the periods with significant linkages. Compared with polarity sentiments (negative vs. positive), detailed sentiments provide more information regarding relationships between Weibo sentiments and the stock market. The stock market exerts positive effects on bullishness and agreement of microblogs. Meanwhile, agreement leads the stock return in-phase at the frequency of approximately 40 trading days, indicating that less disagreement improves certainty about the stock market.

Emoticon-Based Ambivalent Expression: A Hidden Indicator for Unusual Behaviors in Weibo.

Recent decades have witnessed online social media being a big-data window for testifying conventional social theories quantitatively and exploring much detailed human behavioral patterns. In this paper, by tracing the emoticon use in Weibo, a group of hidden "ambivalent users" are disclosed for frequently posting ambivalent tweets containing both positive and negative emotions. Further investigation reveals that this ambivalent expression could be a novel indicator of many unusual social behaviors. For instance, ambivalent users with the female as the majority like to make a sound in midnights and at weekends. They mention their close friends frequently in ambivalent tweets, which attract more replies and serve as a more private communication way. Ambivalent users also respond differently to public affairs from others and demonstrate more interests in entertainment and sports events. Moreover, the sentiment shift in ambivalent tweets is more evident than usual and exhibits a clear "negative to positive" pattern. The above observations, though being promiscuous seemingly, actually point to the self-regulation of negative mood in Weibo, which could find its basis from the traditional emotion management theories in sociology but makes an important extension to the online environment in this study. Finally, as an interesting corollary, ambivalent users are found connected with compulsive buyers and turn out to be perfect targets for online marketing.

Spatio-temporal propagation of cascading overload failures in spatially embedded networks.

Different from the direct contact in epidemics spread, overload failures propagate through hidden functional dependencies. Many studies focused on the critical conditions and catastrophic consequences of cascading failures. However, to understand the network vulnerability and mitigate the cascading overload failures, the knowledge of how the failures propagate in time and space is essential but still missing. Here we study the spatio-temporal propagation behaviour of cascading overload failures analytically and numerically on spatially embedded networks. The cascading overload failures are found to spread radially from the centre of the initial failure with an approximately constant velocity. The propagation velocity decreases with increasing tolerance, and can be well predicted by our theoretical framework with one single correction for all the tolerance values. This propagation velocity is found similar in various model networks and real network structures. Our findings may help to predict the dynamics of cascading overload failures in realistic systems.

Competition between Homophily and Information Entropy Maximization in Social Networks.

In social networks, it is conventionally thought that two individuals with more overlapped friends tend to establish a new friendship, which could be stated as homophily breeding new connections. While the recent hypothesis of maximum information entropy is presented as the possible origin of effective navigation in small-world networks. We find there exists a competition between information entropy maximization and homophily in local structure through both theoretical and experimental analysis. This competition suggests that a newly built relationship between two individuals with more common friends would lead to less information entropy gain for them. We demonstrate that in the evolution of the social network, both of the two assumptions coexist. The rule of maximum information entropy produces weak ties in the network, while the law of homophily makes the network highly clustered locally and the individuals would obtain strong and trust ties. A toy model is also presented to demonstrate the competition and evaluate the roles of different rules in the evolution of real networks. Our findings could shed light on the social network modeling from a new perspective.

Anger is more influential than joy: sentiment correlation in weibo.

Recent years have witnessed the tremendous growth of the online social media. In China, Weibo, a Twitter-like service, has attracted more than 500 million users in less than five years. Connected by online social ties, different users might share similar affective states. We find that the correlation of anger among users is significantly higher than that of joy. While the correlation of sadness is surprisingly low. Moreover, there is a stronger sentiment correlation between a pair of users if they share more interactions. And users with larger number of friends possess more significant sentiment correlation with their neighborhoods. Our findings could provide insights for modeling sentiment influence and propagation in online social networks.

Unraveling the origin of exponential law in intra-urban human mobility.

The vast majority of travel takes place within cities. Recently, new data has become available which allows for the discovery of urban mobility patterns which differ from established results about long distance travel. Specifically, the latest evidence increasingly points to exponential trip length distributions, contrary to the scaling laws observed on larger scales. In this paper, in order to explore the origin of the exponential law, we propose a new model which can predict individual flows in urban areas better. Based on the model, we explain the exponential law of intra-urban mobility as a result of the exponential decrease in average population density in urban areas. Indeed, both empirical and analytical results indicate that the trip length and the population density share the same exponential decaying rate.

Weak ties: subtle role of information diffusion in online social networks.

As a social media, online social networks play a vital role in the social information diffusion. However, due to its unique complexity, the mechanism of the diffusion in online social networks is different from the ones in other types of networks and remains unclear to us. Meanwhile, few works have been done to reveal the coupled dynamics of both the structure and the diffusion of online social networks. To this end, in this paper, we propose a model to investigate how the structure is coupled with the diffusion in online social networks from the view of weak ties. Through numerical experiments on large-scale online social networks, we find that in contrast to some previous research results, selecting weak ties preferentially to republish cannot make the information diffuse quickly, while random selection can achieve this goal. However, when we remove the weak ties gradually, the coverage of the information will drop sharply even in the case of random selection. We also give a reasonable explanation for this by extra analysis and experiments. Finally, we conclude that weak ties play a subtle role in the information diffusion in online social networks. On one hand, they act as bridges to connect isolated local communities together and break through the local trapping of the information. On the other hand, selecting them as preferential paths to republish cannot help the information spread further in the network. As a result, weak ties might be of use in the control of the virus spread and the private information diffusion in real-world applications.