Authorship verification applied to detection of compromised accounts on online social networks

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 %.     

Impact of reference datasets and autocorrelation on classification accuracy

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. 2006 Mannel, 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     

A comparison of the performance of lightweight glass containers manufactured by the P&B and B&B processes

The mechanical performance of lightweight glass packages produced by the NNPB (narrow neck press and blow) process was evaluated by comparison with the same glass packages (regular weight) produced by the conventional process (blow and blow). The temper number (annealing process evaluation), thickness distribution behaviour and mechanical performance (impact, vertical load and thermal shock strength) were analysed before and after line simulation, applied in optimized conditions in the laboratory. The performance of lightweight glass packages under transport simulation (truck envelope) in relation to the regular weight bottles was also evaluated. Both bottles presented residual stress values within the specified limits. The lightweight glass bottles had a more homogenous thickness distribution in comparison with the regular weight bottles and a better performance (about 33% improvement) in relation to the impact strength, especially in the heel, even when evaluated after line simulation. As to the vertical load strength, the lightweight glass bottle also indicated a superior performance to the regular weight bottle, both before and after line simulation. Both bottles withstood the temperature difference of 42°C that such packages are supposed to resist according to thermal shock specification. Due to the better thickness distribution of lightweight glass packages, they withstood a maximum temperature difference (progressive thermal shock) of 5–10°C higher than the regular weight bottles. No restrictions of the lightweight glass packages submitted to the transport simulation were found. Copyright © 2002 John Wiley & Sons, Ltd.     

Gezielte Manipulation von Mikropartikeln in einem Plasma. Manipulation of micro‐disperse particles in a process plasma

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.     

Kinetic Control of Long-Range Cationic Ordering in the Synthesis of Layered Ni-Rich Oxides

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 LiNi_0.6Co_0.2Mn_0.2O₂ (O-LNCM622O) is carefully characterized using in situ synchrotron radiation diffraction. A fast nonequilibrium phase transition from the reactants to a metastable disordered Li_1−x(Ni_0.6Co_0.2Mn_0.2)_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.     

Production and Corrosion Resistance of Thermally Sprayed Fe-Based Amorphous Coatings from Mechanically Milled Feedstock Powders

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 HV_0.3 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 HV_0.3 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.

     

In situ scanning electron microscopy on lithium-ion battery electrodes using an ionic liquid

We present an experimental platform that can be used for investigating lithium-ion batteries with very high spatial resolution. This in situ experiment runs inside a scanning electron microscope (SEM) and is able to track the morphology of an electrode including active and passive materials in real time. In this work it has been used to observe SnO₂ during lithium uptake and release inside a working battery electrode. The experiment strongly relies on an ionic liquid which has very low vapor pressure and can therefore be used as an electrolyte inside the vacuum chamber of the SEM. In contrast to common electrochemical characterization tools, this method allows for the observation of microscopic mechanisms in electrodes. Depending on the SEM, resolutions down to 1 nm can be achieved. As a result, the experimental platform can be used to investigate chemical reaction pathways, to monitor phase changes in electrodes or to investigate degradation effects in batteries. SnO₂ is a potential anode material for future high capacity lithium-ion batteries. Our observations reveal the formation of interface layers, large volume expansions, growth of extrusions, as well as mechanically induced cracks in the electrode particles during cycling.     

Étude des architectures de processeurs parallèles pour le traitement de l’image

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.     

Polyoxometalate Modified Separator for Performance Enhancement of Magnesium–Sulfur Batteries

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.