Compromising legitimate accounts has been the most used strategy to spread malicious content on OSN (Online Social Network). To address this problem, we propose a pure text mining approach to check if an account has been compromised based on its posts content. In the first step, the proposed approach extracts the writing style from the user account. The second step comprehends the k-Nearest Neighbors algorithm (k-NN) to evaluate the post content and identify the user. Finally, Baseline Updating (third step) consists of a continuous updating of the user baseline to support the current trends and seasonality issues of user’s posts. Experiments were carried out using a dataset from Twitter composed by tweets of 1000 users. All the three steps were individually evaluated, and the results show that the developed method is stable and can detect the compromised accounts. An important observation is the Baseline Updating contribution, which leads to an enhancement of accuracy superior of 60 %. Regarding average accuracy, the developed method achieved results over 93 %. 相似文献
Reference data and accuracy assessments via error matrices build the foundation for measuring success of classifications. An error matrix is often based on the traditional holdout method that utilizes only one training/test dataset. If the training/test dataset does not fully represent the variability in a population, accuracy may be over – or under – estimated. Furthermore, reference data may be flawed by spatial errors or autocorrelation that may lead to overoptimistic results. For a forest study we first corrected spatially erroneous ground data and then used aerial photography to sample additional reference data around the field-sampled plots (Mannel et al.2006Mannel, S., Hua, D. and Price, M.2006. A method to obtain large quantities of reference data. International Journal of Remote Sensing, 27: 623–627. [Taylor & Francis Online], [Web of Science ®], [Google Scholar]). These reference data were used to classify forest cover and subsequently determine classification success. Cross-validation randomly separates datasets into several training/test sets and is well documented to perform a more precise accuracy measure than the traditional holdout method. However, random cross-validation of autocorrelated data may overestimate accuracy, which in our case was between 5% and 8% for a 90% confidence interval. In addition, we observed accuracies differing by up to 35% for different land cover classes depending on which training/test datasets were used. The observed discrepancies illustrate the need for paying attention to autocorrelation and utilizing more than one permanent training/test dataset, for example, through a k-fold holdout method.1相似文献
An optimal treatment (surface modification, coating) of micro‐disperse materials (powder, granulate, fibres etc.) by a process plasma requires the understanding of plasma‐particle interaction. For this purpose, related experimental investigations in a novel set‐up PULVA‐INP will be presented. 相似文献
Deciphering the sophisticated interplay between thermodynamics and kinetics of high-temperature lithiation reaction is fundamentally significant for designing and preparing cathode materials. Here, the formation pathway of Ni-rich layered ordered LiNi0.6Co0.2Mn0.2O2 (O-LNCM622O) is carefully characterized using in situ synchrotron radiation diffraction. A fast nonequilibrium phase transition from the reactants to a metastable disordered Li1−x(Ni0.6Co0.2Mn0.2)1+xO2 (D-LNCM622O, 0 < x < 0.95) takes place while lithium/oxygen is incorporated during heating before the generation of the equilibrium phase (O-LNCM622O). The time evolution of the lattice parameters for layered nonstoichiometric D-LNCM622O is well-fitted to a model of first-order disorder-to-order transition. The long-range cation disordering parameter, Li/TM (TM = Ni, Co, Mn) ion exchange, decreases exponentially and finally reaches a steady-state as a function of heating time at selected temperatures. The dominant kinetic pathways revealed here will be instrumental in achieving high-performance cathode materials. Importantly, the O-LNCM622O tends to form the D-LNCM622O with Li/O loss above 850 °C. In situ XRD results exhibit that the long-range cationic (dis)ordering in the Ni-rich cathodes could affect the structural evolution during cycling and thus their electrochemical properties. These insights may open a new avenue for the kinetic control of the synthesis of advanced electrode materials. 相似文献
Mechanically milled FeCrNbB feedstock powders from commercial precursors were used to produce amorphous coatings through two different industrial thermal-spraying techniques: high-velocity oxygen fuel (HVOF) and flame spraying. Microstructure, thermal behavior, and hardness of the coatings and their corrosion resistances in acidic and alkaline chloride-rich media were comparatively studied. HVOF process was effective to produce ~ 200-µm-thick highly amorphous coatings with hardness over than 700 HV0.3 and low porosity (~ 5 pct). Flame-sprayed ~ 220-µm-thick coatings were nanocrystalline, composed of α-Fe, Fe2B, FeNbB, and Fe2O3 phases and presented hardness of 564 HV0.3 and ~ 10 pct porosity. Electrochemical measurements indicated that HVOF coatings exhibit higher corrosion resistance than flame-sprayed ones thanks to the higher amorphous content and lower porosity resulting from the former processing route. Electrochemical impedance spectroscopy results demonstrated that amorphous HVOF Fe60Cr8Nb8B24 (at. pct) coatings are interesting to protect mild steels such as the API 5L X80 against corrosion in chloride-rich environments.
The interest for computervision has led to the proliferation of image processors during the past 25 years. Their study is a challenge, complicating needlessly the access to the domain. This article attempts to solve the problem by presenting a structured analysis of the principal realisations to get a clear understanding of the situation. Then, it becomes clear that most processors are derived from a few basic architectures. An analysis of the trend of developments is presented along with suggestions for futur research. 相似文献
The magnesium–sulfur (Mg-S) battery has attracted considerable attention as a candidate of post-lithium battery systems owing to its high volumetric energy density, safety, and cost effectiveness. However, the known shuttle effect of the soluble polysulfides during charge and discharge leads to a rapid capacity fade and hinders the realization of sulfur-based battery technology. Along with the approaches for cathode design and electrolyte formulation, functionalization of separators can be employed to suppress the polysulfide shuttle. In this study, a glass fiber separator coated with decavanadate-based polyoxometalate (POM) clusters/carbon composite is fabricated by electrospinning technique and its impacts on battery performance and suppression of polysulfide shuttling are investigated. Mg–S batteries with such coated separators and non-corrosive Mg[B(hfip)4]2 electrolyte show significantly enhanced reversible capacity and cycling stability. Functional modification of separator provides a promising approach for improving metal–sulfur batteries. 相似文献