Proactive handling of social unrest events which are common happenings in both democracies and authoritarian regimes requires that the risk upcoming event is continuously assessed. Most existing approaches comparatively pay little attention to considering development stages. In this paper, we use autocoded dataset GDELT (Global Data on Events, Location, Tone) build a Hidden Markov Models (HMMs) based framework predict indicators associated with country instability. The utilizes temporal...

A radial line slot antenna (RLSA) is a nonresonant multimode waveguide slotted array. To improve the power handing capacity, we design slow wave structure which replaces traditional dielectric sheet. At same time, of radiator unit redesigned and optimised. This high-power microwave (HPM) RLSA fed from double-layered to realize directional radiation microwave. However, slots spirally arrayed for circular polarization give rise serious reflection in higher order modes, disturbs normal...

Twitter trends, a timely updated set of top terms in Twitter, have the ability to affect public agenda community and attracted much attention. Unfortunately, wrong hands, trends can also be abused mislead people. In this paper, we attempt investigate whether are secure from manipulation malicious users. We collect more than 69 million tweets 5 accounts. Using collected tweets, first conduct data analysis discover evidence trend manipulation. Then, study at topic level infer key factors that...

Frequent subgraph mining (FSM) plays an important role in graph mining, attracting a great deal of attention many areas, such as bioinformatics, web data and social networks. In this paper, we propose SSiGraM (Spark based Single Graph Mining), Spark parallel frequent algorithm single large graph. Aiming to approach the two computational challenges FSM, conduct extension support evaluation across all distributed cluster worker nodes. addition, also employ heuristic search strategy three novel...

Machine reading comprehension (MRC) has gained increasingly wide attention over the past few years. A variety of benchmark datasets have been released, which triggers development quite a MRC approaches based on deep learning techniques. However, most existing models are designed to address English MRC. When applying them directly Chinese documents, performance often degrades considerably because some special characteristics Chinese, inevitable segmentation errors in particular. In this...

Due to the existence of information overload in social networks, it becomes increasingly difficult for users find useful according their interests. This paper takes Twitter-like networks into account and proposes models characterize process diffusion under overload. Users are classified different types in-degrees out-degrees, user behaviors generalized two categories: generating forwarding. View scope is introduced model information-processing capability overload, average number times a...

The opinion spreading process can be modeled as the spread of an epidemic through a network, which assumes homogeneous relationships between individuals. However, positive and negative in signed networks play different roles process, following general rule that same will diffuse friends, while opposite likely emerge out interactions enemies. In order to explore behavior networks, we proposed simple model based on susceptible-infected-recovered (SIR) model. Under assumption mixing, also...

Sina-Microblog, the earliest and biggest microblogging service in China, has become one of most popular media information propagation. In order to gain insights into topological diffusing characteristics network we crawled Sina-Microblog for about 3 months obtain trace its topology topics. Compared with other online social networks, our measurement study shows a number interesting findings. Our data suggests that apparent small-world effect scale-free characteristic, specially, outdegree...

We introduce a computational approach to detecting and predicting occupy protest events over daily timeframe. Existing methods mainly solve this problem by document clustering techniques. This paper proposes novel graph-based event detection framework which applies subgraph pattern mining for task. In addition, we also use standard linear logistic regression model estimate the probability of an OPE occurring on specific day. A wealth data about Occupy Wall Street in New York Central Hong...

Twitter as an online social network is used for many reasons, including information dissemination, marketing, political organizing, spamming, promotion, conversations and so on. Characterizing these activities categorizing users a challenging task. Traditional user classification models are based on individual user's profile such age, location, register time, interests tweets, which have not considered the whole complexity of posting behavior. In this paper we introduce Multi-scale Entropy...

The influence of fibrous velvet cathodes on the electrical stability a planar high-power diode powered by ∼230 kV, ∼110 ns pulse has been investigated. current density was order ∼123 A/cm2. A combination time-resolved and optical diagnostics employed to study basic phenomenology temporal spatial evolution plasmas. Additionally, an impedance model used extract information about this plasma from voltage profiles. results two were compared. By comparison with commercial polymer cathode, dense...

Social networks contain a large amount of information on transportation, e.g., traffic accidents, congestions, and vehicles. Such is the original ideas people with respect to real-world transportation issues, detecting communities topic from will benefit many ITS applications. However, realworld social network nodes often multiple attributes, can be very large. The two properties lead confusion unscalability problem clustering methods. In this paper, we propose semisupervised method, namely,...

Social networks have attracted remarkable attention and it is of great importance to understand the process opinion spreading in popular social networks. However, most research on diffusion cannot be applied directly investigate networks, where relationships are heterogeneous structural balance a common phenomenon. In this paper, we propose models characterize signed under impact balance. We classify users into different types according numbers their positive links, define term user...

The Big data analytics gives new chances to the enterprises enhance their management and manufacturing levels. A solution with case study is proposed accomplish deep-level quality based on big analytics. First, implementation of industrial process illustrated illustration. Through analysis feature extraction off-line data, corresponding reference model library constructed, which can be used for real-time processing unlabeled in field. cluster, outlier other indicators we get from calculation...

A new genre of P2P systems, Video-on-Demand (P2P-VoD), has not only received substantial recent research attention, but also been implemented and deployed with success in large-scale real-world streaming such as PPLive. The essential characterization P2P-VoD systems is to take full advantage peer upload bandwidth contribution a cache on each peer. Meanwhile, the asynchronous playback dilutes peers' ability assisting other. How can peers assist other systems? In this paper, we seek develop...

Social networks have attracted remarkable attention from both academic and industrial societies it is of great importance to understand the formation social networks. However, most existing research cannot be applied directly investigate networks, where relationships are heterogeneous structural balance a common phenomenon. In this paper, we take positive negative into consideration propose model characterize process network under impact balance. model, new node first establishes link with...

Recent empirical studies have discovered that many social networks heterogeneous relationships, which are signed and weighted relationships between individual nodes. To explore the pattern of opinion dynamics in diverse with we set up a general agent-based simulation framework named interaction network (OIN), propose novel model dynamics, influence agents depends on their relationships. Then, by conducting series simulations based OIN, find opinions at steady state depend degree...

In this paper, we conduct a comparative study on two different text features in corpus clustering: TF*IDF feature and Topic feature. The former is mainly used similarity-based clustering methods, while the latter, which produced by LDA model, to identify topics of texts. We experiments 20-newsgroups (20NG) datasets. Based dataset, typical methods are respectively employed compare performance above features. demonstrate if optimal topic number chosen, outperforms accuracy.

A behavioral model describing the nonlinearity of a double-balanced mixer is presented. The analysis based on diode I-V characteristics. double frequency test method used to extract mixer. This can measure main channel and all spurious response. measurement carried out in commercial mixer, results show that nonlinear

In this paper an improved virtual chamber for in-situ radiated emissions measurement is developed. Virtual test system a method electromagnetic interference (EMI) which based on suppressing ambient signals from signal containing of the equipment under (EUT) and circumstance. Nevertheless, technique assumption that are either time asymmetrical or space asymmetrical. Thus, saltation in time. Moreover, not symmetrical reality. We present combined sharing synchronized measurement. This can...

In complex network epidemiology, diseases spread along contacting edges between individuals and different may play roles in epidemic outbreaks. Quantifying the efficiency of is an important step towards arresting epidemics. this paper, we study general susceptible-infected-recovered models, introduce transmission capability to measure edges. Results show that deleting with highest will greatly decrease epidemics on scale-free networks. Basing message passing approach, get exact mathematical...

As pre-trained models, like Transformers, are increasingly deployed on cloud platforms for inference services, the privacy concerns surrounding model parameters and data becoming more acute. Current Privacy-Preserving Transformer Inference (PPTI) frameworks struggle with "impossible trinity" of privacy, efficiency, performance. For instance, Secure Multi-Party Computation (SMPC)-based solutions offer strong guarantees but come significant overhead performance trade-offs. On other hand, PPTI...