Rising demand for ruminant meat and dairy products in developing countries is expected to double anthropogenic greenhouse gas and ammonia emissions from livestock by 2050. Mitigation strategies are urgently needed to meet demand while minimizing environmental impacts. Here, we develop scenarios for mitigating emissions under local vs global supply policies using data from 308 livestock farms across mainland China, where emissions intensities are ~50% higher than those in developed nations. Intensification of domestic production and globalized expansion through increased trade result in reductions in global emissions by nearly 30% over a business-as-usual scenario, but at the expense of trading partners absorbing the associated negative externalities of environmental degradation. Only adoption of a mixed strategy combining global best-practice in sustainable intensification of domestic production, with increased green-source trading as a short-term coping strategy, can meet 2050 demand while minimizing the local and global environmental footprint of China’s ruminant consumption boom.
The filter bubble is an intermediate structure to provoke polarization and echo chambers in social networks, and it has become one of today's most urgent issues for social media. Previous studies usually equated filter bubbles with community structures and emphasized this exogenous isolation effect, but there is a lack of full discussion of the internal organization of filter bubbles. Here, we design an experiment for analysing filter bubbles taking advantage of social bots. We deployed 128 bots to Weibo (the largest microblogging network in China), and each bot consumed a specific topic (entertainment or sci-tech) and ran for at least two months. In total, we recorded about 1.3 million messages exposed to these bots and their social networks. By analysing the text received by the bots and motifs in their social networks, we found that a filter bubble is not only a dense community of users with the same preferences but also presents an endogenetic unidirectional star-like structure. The structure could spontaneously exclude non-preferred information and cause polarization. Moreover, our work proved that the felicitous use of artificial intelligence technology could provide a useful experimental approach that combines privacy protection and controllability in studying social media.
Recommendation algorithms are data filtering tools that make use of algorithms and data to recommend the most relevant items to a particular user. The algorithm-driven recommenders become indispensable and supersede search engines as the most important information dissemination channel. On one hand, it becomes an integral component in the existing social media, e.g. Weibo, Twitter, etc. On the other hand, news aggregators and recommenders have proliferated and gained an increasing market share. As a result, the previous studies usually study the “filter bubbles” phenomenon in the context where the social filtering dominates the dissemination of information. However, less attention is paid to the news aggregators and recommenders where algorithm-driven technological filtering dominates. Therefore, in the previous research, “filter bubbles” are usually equated with the community structure, but lack of the detailed analysis of the content agglomeration through the users’ interaction with the platforms. Based on these concerns, we propose a four-phase (“Selection”, “Setup”, “Link”, and “Evaluation”) skeletal solution framework targeted at exploiting the filter bubble effect of the personalized news aggregation and recommendation system. Furthermore, we illustrate the effectiveness of the proposed framework with a case study in three top Chinese news aggregators, i.e. Toutiao, Baidu News, and Tencent News. The results show that the users are narrowed into one or a limited number of topics over time. The phenomenon of the narrowed topics is deemed as the emergence of the “filter bubbles”. We also observe that the filter bubbles demonstrate different convergence degrees as user’s individual preference varies.
Biodiversity has a close relationship with ecosystem functioning. For most biodiversity–ecosystem‐functioning studies, biodiversity has been linked to a single indicator variable of ecosystem functioning. However, there are generally multiple ecosystem processes contributing to ecosystem functioning and they differ in their dependence on biodiversity. Thus, the relationship between biodiversity and ecosystem functioning can be stronger when multiple rather than single ecosystem processes are considered. Using both mass‐balance and stable‐isotope approaches, we explored the effects of plant diversity on nitrogen (N) removal sustained by multiple N‐cycling processes in experimental microcosms simulating constructed wetlands, an ecosystem treating wastewater with high N loading. Four species were used to assemble different plant communities, ranging in richness from one to four species. The removal of N, indicated by low levels of total inorganic N concentration (TIN) present in the effluent, was considered as an integrated measure of ecosystem functioning, combining three constituent N‐cycling processes: plant uptake, denitrification, and substrate adsorption. Our results showed that (1) species richness had a positive effect on N removal, in particular, the four‐species mixture reduced effluent TIN to a lower level than any monoculture; however, polycultures (two‐, three‐, and four‐species mixtures) did not outperform the most efficient monoculture when each of the three constituent N‐cycling processes was considered by itself; (2) species identity had significant impacts on single processes. Communities with the speciesCoix lacryma‐jobishowed the greatest capacity for N uptake and communities withPhragmites australishad the highest denitrification rates; (3) isotope fractionation in the rhizosphere ofCoix lacryma‐jobiwas primarily due to microbial denitrification while multistep isotope fractionation was detected forPhragmites australisandAcorus calamus(indicating recycling of N), suggesting that species differed in the way they transformed N; (4) the enhanced N removal at high diversity may be due to mutualistic interactions among species belonging to different functional types. Our findings demonstrated that although plant species richness had negligible effects on individual N‐cycling processes, it enhanced the overall ecosystem functioning (N removal) when these processes were considered collectively. Our study thus contributes to improve the treatment efficiency of constructed wetlands through proper vegetation management.
27 August 2021 in IEEE Transactions on Computational Social Systems
Location-Based Social Networks have been widely studied in recent years; new approaches constantly developed to solve individuals’ trajectory prediction tasks. However, most of these methods require sufficient data to learn individual features, which is not always satisfied in real situations, especially for online data. The digital data on human behavior typically follows a power-law distribution, indicating that only a few people have rich activities recorded while most people’s behavioral data are limited. In order to overcome this hurdle, our work constructs the user behavior proximity network (UBPN) and proposes a new walking strategy based on this network that extracts the hidden information from the social contacts to substitute the unobserved behavioral information of an individual. Specifically, our proposed walking strategy has two walking paths, accounting for the temporal and social information on the ego users’ and their alters’ mobility activities. This walking strategy is model-agnostic and can be integrated with many existing walk-based deep learning methods. Our work applies the methods on two real-world datasets with rich spatiotemporal information and shows that the performances of the existing prediction methods improve significantly by integrating the proposed walking strategy.
In recent years, fake news detection and its characteristics have attracted a number of researchers. However, most detection algorithms are driven by data rather than theories, which causes the existing approaches to only perform well on specific datasets. To the extreme, several features only perform well on specific datasets. In this study, we first define the feature drift in fake news detection methods, and then demonstrate the existence of feature drift and use interpretable models (i.e., Shapley Additive Explanations and Partial Dependency Plots) to verify the feature drift. Furthermore, by controlling the distribution of tweets’ creation times, a novel sampling method is proposed to explain the reason for feature drift. Finally, the Anchors method is used in this paper as a supplementary interpretation to exhibit the potential characteristics of feature drift further. Our work provides deep insights into the temporal patterns of fake news detection, proving that the model’s performance is also highly related to the distribution of datasets.