Geo-tagged tweets can potentially help with sensing the interaction of people with their surrounding environment. Based on this hypothesis, this paper makes use of geotagged tweets in order to ascertain various land uses with a broader goal to help with urban/city planning. The proposed method utilises supervised learning to reveal spatial land use within cities with the help of Twitter activity signatures. Specifically, the technique involves using tweets from three cities of Australia namely Brisbane, Melbourne and Sydney. Analytical results are checked against the zoning data provided by respective city councils and a good match is observed between the predicted land use and existing land zoning by the city councils. We show that geo-tagged tweets contain features that can be useful for land use identification.

Federated learning (FL) is a promising distributed learning solution that only exchanges model parameters without revealing raw data. However, the centralized architecture of FL is vulnerable to the single point of failure. In addition, FL does not examine the legitimacy of local models, so even a small fraction of malicious devices can disrupt global training. To resolve these robustness issues of FL, in this paper, we propose a blockchain-based decentralized FL framework, termed VBFL, by exploiting two mechanisms in a blockchained architecture. First, we introduced a novel decentralized validation mechanism such that the legitimacy of local model updates is examined by individual validators. Second, we designed a dedicated proof-of-stake consensus mechanism where stake is more frequently rewarded to honest devices, which protects the legitimate local model updates by increasing their chances of dictating the blocks appended to the blockchain. Together, these solutions promote more federation within legitimate devices, enabling robust FL. Our emulation results of the MNIST classification corroborate that with 15% of malicious devices, VBFL achieves 87% accuracy, which is 7.4x higher than Vanilla FL.

After affecting the world in unexpected ways, COVID-19 has started mutating which is evident with the insurgence of its new variants. The governments, hospitals, schools, industries, and humans, in general, are looking for a potential solution in the vaccine which will eventually be available but its timeline for eradicating the virus is yet unknown. Several researchers have encouraged and recommended the use of good practices such as physical healthcare monitoring, immunity-boosting, personal hygiene, mental healthcare, and contact tracing for slowing down the spread of the virus. In this article, we propose the use of wearable/mobile sensors integrated with the Internet of Everything to cover the spectrum of good practices in an automated manner. We present hypothetical frameworks for each of the good practice modules and propose the COvid-19 Resistance Framework using the Internet of Everything (CORFIE) to tie all the individual modules in a unified architecture. We envision that CORFIE would be influential in assisting people with the new normal for current and future pandemics as well as instrumental in halting the economic losses, respectively. We also provide potential challenges and their probable solutions in compliance with the proposed CORFIE.

Continual relation extraction is an important task that focuses on extracting new facts incrementally from unstructured text. Given the sequential arrival order of the relations, this task is prone to two serious challenges, namely catastrophic forgetting and order-sensitivity. We propose a novel curriculum-meta learning method to tackle the above two challenges in continual relation extraction. We combine meta learning and curriculum learning to quickly adapt model parameters to a new task and to reduce interference of previously seen tasks on the current task. We design a novel relation representation learning method through the distribution of domain and range types of relations. Such representations are utilized to quantify the difficulty of tasks for the construction of curricula. Moreover, we also present novel difficulty-based metrics to quantitatively measure the extent of order-sensitivity of a given model, suggesting new ways to evaluate model robustness. Our comprehensive experiments on three benchmark datasets show that our proposed method outperforms the state-of-the-art techniques. The code is available at the anonymous GitHub repository: https://github.com/wutong8023/AAAI_CML.

Physiological functions of the microtubule cytoskeleton are expected to be regulated by a variety of posttranslational tubulin modifications. For instance, tubulin glycylation is almost exclusively found in cilia and flagella, but its role in the function of these organelles remains unclear. Gadadhar et al. now demonstrate in mice that glycylation, although nonessential for the formation of cilia and flagella, coordinates the beat waveform of sperm flagella. This activity is a prerequisite for progressive sperm swimming and thus for male fertility. At the ultrastructural level, lack of glycylation perturbed the distribution of axonemal dynein conformations, which may explain the observed defects in flagellar beat. Science , this issue p. [eabd4914][1] ### INTRODUCTION Microtubules are key components of the eukaryotic cytoskeleton. Although they are involved in a wide variety of functions, microtubules are structurally highly similar across most cell types and organisms. It was suggested that a “tubulin code,” formed by combinations of tubulin posttranslational modifications, adapts individual microtubules to specific functions within living cells. However, clear-cut functional and mechanistic data verifying this concept are still scarce. Glycylation is among the least explored posttranslational modifications of tubulin and has, so far, exclusively been found on microtubules of cilia and flagella from a variety of species. Previous work has suggested that glycylation might be essential for cilia and flagella, but mechanistic insight remains lacking. ### RATIONALE Two enzymes from the tubulin-tyrosine ligase-like (TTLL) family, TTLL3 and TTLL8, are essential to initiate glycylation of tubulin in mammals. To entirely abolish glycylation at the organism level and to determine its physiological function, we generated a double-knockout mouse lacking both glycylating enzymes ( Ttll3−/−Ttll8−/− ). Inactivation of these two enzymes led to a lack of glycylation in all analyzed cilia and flagella. This allowed us to investigate the role of glycylation in the function of these organelles. ### RESULTS Despite the absence of glycylation in Ttll3−/−Ttll8−/− mice, no gross defects were observed at the organism and tissue levels. Motile ependymal cilia in brain ventricles as well as motile cilia in the respiratory tract were present and appeared normal. Sperm flagella were also assembled normally, and sperm were able to swim. However, in vitro fertility assays showed that male Ttll3−/−Ttll8−/− mice were subfertile. Computer-assisted sperm analyses revealed motility defects of Ttll3−/−Ttll8−/− sperm. Further analyses showed that lack of glycylation leads to perturbed flagellar beat patterns, causing Ttll3−/−Ttll8−/− sperm to swim predominantly along circular paths. This is highly unusual for mammalian sperm and interferes with their ability to reach the oocyte for fertilization. To determine the molecular mechanisms underlying this aberrant flagellar beat, we used cryo–electron tomography. The three-dimensional structure of the 96-nm repeat of the Ttll3−/−Ttll8−/− sperm axoneme showed no aberrations in its overall assembly. By contrast, the structure of both outer and inner dynein arms (ODAs and IDAs) was perturbed in Ttll3−/−Ttll8−/− flagella. Classification analysis showed that the incidence and distribution of pre-powerstroke and post-powerstroke conformations of ODAs and IDAs were altered in Ttll3−/−Ttll8−/− sperm. These ultrastructural findings indicate that glycylation is required to efficiently control the dynein powerstroke cycle, which is essential for the generation of a physiological flagellar beat. ### CONCLUSION Our work shows that tubulin glycylation regulates the beat of mammalian flagella by modulating axonemal dynein motor activity. Lack of glycylation leads to perturbed sperm motility and male subfertility in mice. Considering that human sperm are more susceptible than mouse sperm to deficiencies in sperm motility, our findings imply that a perturbation of tubulin glycylation could underlie some forms of male infertility in humans. ![Figure][2] Tubulin glycylation controls sperm motility. ( A ) Microtubules in sperm flagella are rich in tubulin posttranslational modifications. Mice deficient for the glycylating enzymes TTLL3 and TTLL8 lack glycylation. ( B ) Mammalian sperm swim in linear paths. In the absence of glycylation, abnormal, mostly circular swimming patterns are observed, which impede progressive swimming. ( C ) Absence of glycylation leads to perturbed distribution of axonemal dynein conformations in Ttll3−/−Ttll8−/− flagella, which impedes normal flagellar beating. Posttranslational modifications of the microtubule cytoskeleton have emerged as key regulators of cellular functions, and their perturbations have been linked to a growing number of human pathologies. Tubulin glycylation modifies microtubules specifically in cilia and flagella, but its functional and mechanistic roles remain unclear. In this study, we generated a mouse model entirely lacking tubulin glycylation. Male mice were subfertile owing to aberrant beat patterns of their sperm flagella, which impeded the straight swimming of sperm cells. Using cryo–electron tomography, we showed that lack of glycylation caused abnormal conformations of the dynein arms within sperm axonemes, providing the structural basis for the observed dysfunction. Our findings reveal the importance of microtubule glycylation for controlled flagellar beating, directional sperm swimming, and male fertility. [1]: /lookup/doi/10.1126/science.abd4914 [2]: pending:yes

Cyber-physical systems (CPS) and Internet-of-Things (IoT) devices are increasingly being deployed across multiple functionalities, ranging from healthcare devices and wearables to critical infrastructures, e.g., nuclear power plants, autonomous vehicles, smart cities, and smart homes. These devices are inherently not secure across their comprehensive software, hardware, and network stacks, thus presenting a large attack surface that can be exploited by hackers. In this article, we present an innovative technique for detecting unknown system vulnerabilities, managing these vulnerabilities, and improving incident response when such vulnerabilities are exploited. The novelty of this approach lies in extracting intelligence from known real-world CPS/IoT attacks, representing them in the form of regular expressions, and employing machine learning (ML) techniques on this ensemble of regular expressions to generate new attack vectors and security vulnerabilities. Our results show that 10 new attack vectors and 122 new vulnerability exploits can be successfully generated that have the potential to exploit a CPS or an IoT ecosystem. The ML methodology achieves an accuracy of 97.4% and enables us to predict these attacks efficiently with an 87.2% reduction in the search space. We demonstrate the application of our method to the hacking of the in-vehicle network of a connected car. To defend against the known attacks and possible novel exploits, we discuss a defense-in-depth mechanism for various classes of attacks and the classification of data targeted by such attacks. This defense mechanism optimizes the cost of security measures based on the sensitivity of the protected resource, thus incentivizing its adoption in real-world CPS/IoT by cybersecurity practitioners.

We release an open library, called TextBox, which provides a unified, modularized, and extensible text generation framework. TextBox aims to support a broad set of text generation tasks and models. In TextBox, we implements several text generation models on benchmark datasets, covering the categories of VAE, GAN, pre-trained language models, etc. Meanwhile, our library maintains sufficient modularity and extensibility by properly decomposing the model architecture, inference, learning process into highly reusable modules, which allows easily incorporating new models into our framework. It is specially suitable for researchers and practitioners to efficiently reproduce baseline models and develop new models. TextBox is implemented based on PyTorch, and released under Apache License 2.0 at https://github.com/RUCAIBox/TextBox.

Artificial Intelligence (AI) is being increasingly deployed in practical applications. However, there is a major concern whether AI systems will be trusted by humans. In order to establish trust in AI systems, there is a need for users to understand the reasoning behind their solutions. Therefore, systems should be able to explain and justify their output. In this paper, we propose an argument scheme-based approach to provide explanations in the domain of AI planning. We present novel argument schemes to create arguments that explain a plan and its key elements; and a set of critical questions that allow interaction between the arguments and enable the user to obtain further information regarding the key elements of the plan. Furthermore, we present a novel dialogue system using the argument schemes and critical questions for providing interactive dialectical explanations.

Working with a non-stationary stream of data requires for the analysis system to evolve its model (the parameters as well as the structure) over time. In particular, concept drifts can occur, which makes it necessary to forget knowledge that has become obsolete. However, the forgetting is subjected to the stability-plasticity dilemma, that is, increasing forgetting improve reactivity of adapting to the new data while reducing the robustness of the system. Based on a set of inference rules, Evolving Fuzzy Systems-EFS-have proven to be effective in solving the data stream learning problem. However tackling the stability-plasticity dilemma is still an open question. This paper proposes a coherent method to integrate forgetting in Evolving Fuzzy System, based on the recently introduced notion of concept drift anticipation. The forgetting is applied with two methods: an exponential forgetting of the premise part and a deferred directional forgetting of the conclusion part of EFS to preserve the coherence between both parts. The originality of the approach consists in applying the forgetting only in the anticipation module and in keeping the EFS (called principal system) learned without any forgetting. Then, when a drift is detected in the stream, a selection mechanism is proposed to replace the obsolete parameters of the principal system with more suitable parameters of the anticipation module. An evaluation of the proposed methods is carried out on benchmark online datasets, with a comparison with state-of-the-art online classifiers (Learn++.NSE, PENsemble, pclass) as well as with the original system using different forgetting strategies.

In this paper, a novel training paradigm inspired by quantum computation is proposed for deep reinforcement learning (DRL) with experience replay. In contrast to traditional experience replay mechanism in DRL, the proposed deep reinforcement learning with quantum-inspired experience replay (DRL-QER) adaptively chooses experiences from the replay buffer according to the complexity and the replayed times of each experience (also called transition), to achieve a balance between exploration and exploitation. In DRL-QER, transitions are first formulated in quantum representations, and then the preparation operation and the depreciation operation are performed on the transitions. In this progress, the preparation operation reflects the relationship between the temporal difference errors (TD-errors) and the importance of the experiences, while the depreciation operation is taken into account to ensure the diversity of the transitions. The experimental results on Atari 2600 games show that DRL-QER outperforms state-of-the-art algorithms such as DRL-PER and DCRL on most of these games with improved training efficiency, and is also applicable to such memory-based DRL approaches as double network and dueling network.