2026-03-25
Mouhssine El Atillah
Due to its complexity, the Arabic language and its extensions build a fertile field of research in the field of artificial intelligence in general and optical character recognition (OCR) specifically. There are several languages that use the Arabic alphabet in their manuscripts. These languages innovated new letters to pronounce sounds not found in the Arabic language. These letters are called 'Arabic-derived letters'. To enrich the Arabic language, we can use these letters to know the true pronunciation of intrusive words in the Arabic language. This article deals with the Arabic-derived letters (ADL) dataset. It is a new dataset that consists of 55440 scanned images of papers written by 30 participants of different ages, with a data augmentation technique to increase the number of images. This study aims to evaluate and compare the effectiveness of different convolutional neural network architectures for ADL recognition, focusing on accuracy, robustness, and generalization capability. Three architectures were implemented: LeNet, a simplified ResNet model with residual blocks, and a deep VGG-Like network. Training was limited to 40 epochs with early stopping after 5 epochs without improvement. Experimental results show that the VGG-Like model achieves the best performance with 99.61% accuracy in validation, closely followed by ResNet with an accuracy of 98.98%. In contrast, LeNet performs less efficiently by 96.43%. These results clearly demonstrate that modern and deep architectures provide better accuracy and robustness for the classification of handwritten characters.
2026-03-25
Ahmed L. Alshami et al.
Botnets turned into a security problem that would put user privacy and security at risk. Progressive and flexible Machine Learning (ML) techniques are necessary for robust botnet revealing. In this study researchers presented an integration of the modified Naïve Bayes and M3L algorithms that uses Factor Analysis of Mixed Data (FAMD) for feature aggregation, and Laplace smoothing for adjustment to address the limitation of conventional Naïve Bayes classifiers for detecting botnet. The modified algorithm uses Laplace smoothing to address problems with zero-frequency data and improve the classifier's adaptability. FAMD was used to merge continuous and categorical features to improve the algorithm's capacity and handle mixed data types and lowering the feature dimensionality. Better classification performance and less computing complexity follow from this. Through using a dataset of network traffic that includes both benign and botnet, the proposed approach was evaluated. The suggested algorithm achieves a 97.45% accuracy and a Mean Squared Error (MSE) of 0.039. These results show the ability of the combined method in accurately detect botnet activity and reduce related security threats.
2026-03-25
Saad Naji Al-Azzawi et al.
This paper introduces a new nonclassical statistical distribution, SMART1, derived from a nonpolynomial function. This function is used in ecology to model population growth rates, in medicine to classify the relationship between activity and tumor volume in cancer, and in economics to analyze the relationship between supply and demand. The proposed distribution differs from the Gompertz distribution, which is fundamentally based on the exponential function. By contrast, the SMART1 distribution is constructed on principles of mathematical analysis, specifically through the identification of local maximum endpoints of the Gompertz growth function and the subsequent verification that the resulting function satisfies the criteria of a probability density function. This distribution (SMART1) is flexible and can model various real-world phenomena on a bounded interval ( 0, β ), especially in reliability analysis or survival modeling, where an upper lifetime limit exists. All statistical concepts are expressed in terms of the distribution's scale parameter β defines the upper bound (or support limit) of the distribution. The random variable X can only take values in the interval (0, β ). In other words, β is a ``lifetime limit'' or ``maximum capacity''; and the shape parameter α determines how the risk or likelihood is distributed over time: Low α : high initial risk that decreases (e.g., early failures). High α : low initial risk that increases with time (e.g., aging or wear-out). These include the probability density function, the cumulative distribution function, the reliability function, the hazard function, order statistics, moments, and key measures such as the mode and the median.
2026-03-25
Malak Alnimer et al.
Let $\mathfrakB$ be a graded commutative ring with unity, and let $\mathfrakW$ be a graded unital $\mathfrakB$-module. This study introduces and develops the concept of graded strongly $J_gr^Soc$-2-absorbing submodules, a natural extension of graded $J_gr$-2-absorbing submodules within the framework of graded module theory. The motivation for this generalization stems from the need to better capture the interplay between graded algebraic structures and the behaviors of certain radicals and socles under graded operations. A properly graded submodule $N$ of $\mathfrakW$ is defined as a graded strongly $J_gr^Soc$-2-absorbing submodule if, for all $b,uin h( \mathfrakB )$ and $cin h( \mathfrakW )$, the containment $bucin N$ implies that at least one of the following conditions holds: $bcin N + ( J_gr( \mathfrakW ) \cap Soc^gr( \mathfrakW ) )$, $ucin N + ( J_gr( \mathfrakW ) \cap Soc^gr( \mathfrakW ) )$, or $buin ( N + ( J_gr( \mathfrakW ) \cap Soc^gr( \mathfrakW ) ):_\mathfrakB\mathfrakW )$. Several fundamental properties of these submodules are established, along with characterizations that distinguish them from related graded structures. Moreover, the investigation reveals meaningful connections between these submodules and the graded socle and graded Jacobson radical of the module, offering new insights into their algebraic significance.
2026-03-23
Rajaa K. Mohammad et al.
Density Functional Theory (DFT) was employed to investigate the effects of beryllium (Be) substitution on the vibrational and electronic properties of a single-wall carbon nano-cone. The study aims to clarify how Be incorporation modifies the structural stability and electronic characteristics of this nanostructure. Electronic parameters including ionization potential (I), electron affinity (Eea), Fermi energy (Ef), HOMO–LUMO energies, total energy (E), and band gap (Eg) were calculated. Infrared (IR) spectra, optimized geometries, electrostatic potential maps, and electron density distributions were also analyzed. The results indicate that Be substitution reduces the ionization potential while increasing the electron affinity. Incorporation of Be atoms raises the HOMO level, Fermi energy, electron affinity, and total energy, whereas it lowers the LUMO energy and narrows the band gap. Vibrational analysis shows weakened vibrational stability upon substitution. Furthermore, electrostatic potential and charge density distributions are strongly influenced by the type, number, and position of substituted atoms, as well as by local charge redistribution. Positional dependence of Be substitution leads to noticeable variations in both electronic and vibrational responses. These findings demonstrate that controlled Be substitution can effectively tune the electronic structure of carbon nano-cones, enhancing their suitability for potential applications in nanoelectronics, catalysis, sensing, and energy storage systems.
2026-03-23
Ari Sulistyo Rini et al.
In this work, bio-colloidal silver nanoparticles (AgNPs) were synthesized using Matoa ( Pometia pinnata ) leaf extract as a reducing and stabilizing agent under microwave irradiation to promote rapid AgNPs formation. The influence of microwave irradiation powers (360, 540, and 720 W) on the optical, structural, and morphological properties of the synthesized AgNPs was systematically investigated. The AgNPs formation was characterized by yellowish-brown colloids formation with surface plasmon resonant (SPR) bands centered around 430 nm, as detected by UV-Vis spectroscopy. FTIR analysis revealed the presence of C-H and O-H functional groups from polyphenols and flavonoids in Ag + ion reduction. XRD pattern revealed a face-centered cubic (FCC) crystalline structure of AgNPs, while TEM images showed predominantly quasi-spherical morphology with average diameter of 58–62 nm. The colorimetric sensing performance was examined over Hg 2+ concentration ranging from 1–100 ppm. Upon Hg 2+ exposure, the AgNPs colloids demonstrated a progressive decrease in SPR intensity along with visible discoloration in the 350–550 nm region. Nanoparticles synthesized at 720 W exhibited the highest sensitivity of AgNPs, which displayed obvious discoloration observed in the range 40–100 ppm of Hg 2+ with steepest slope at 8 × 10 –4 . These findings highlight the potential of microwave-assisted Pometia pinnata -derived AgNPs as a green and eco-friendly approach for the colorimetric sensing of mercury ions.
2026-03-23
Muzhda Qasim Qader et al.
Background: This study evaluated the seasonal and spatial distribution of twelve heavy metals in the groundwater samples collected in fifty wells with the aim of assessing the quality of water and health hazards associated with it. The data were analyzed to compare the wet and dry seasons were the analytical data using pollution indices, the Nemerow Pollution Index (NPI), and the Single-Factor Pollution Index (SFPI). The results revealed that the water quality was mainly excellent (NPI