Journal Of AL-Turath University College

Cyber threats pose an increasing risk to organizations due to the growing complexity and frequency
of attacks. Traditional security systems often fail to detect advanced threats, leading to the need for
intelligent and automated solutions. This research proposes an intelligent cyber threat detection and
sharing system based on Artificial Neural Networks (ANNs), using the CICIDS dataset to classify
and identify cyber-attacks such as DDoS and infiltration. The system integrates data preprocessing,
model training, evaluation, and a RESTful API for secure threat intelligence sharing. The proposed
ANN model achieved a classification accuracy of 94%, precision of 93%, and recall of 95%.
Additionally, a collaborative framework and feedback mechanism were implemented to enhance
inter-organizational security cooperation. This study demonstrates the feasibility of ANN-based
intelligent systems for proactive cybersecurity and establishes a foundation for continuous learning
and secure information exchange.

Image processing is vital in many areas of human endeavor today. Images often suffer from poor
illumination due to poor lighting, poor backlighting, or other conditions, resulting in low-light
images, which are difficult to process in computer vision and human recognition. Therefore,
processing to improve these low-light images becomes important in computer vision applications.
This work aims to use the Retinex algorithm to improve low-light images through a two-stage
process. First, dynamic range scaling is applied to each input channel, followed by transforming
the result to the frequency domain using a fast Fourier transform (FFT). Multiscale Retinex is
applied in the frequency domain before returning the image to the spatial domain. The second stage
consists of improving the visual appearance of details by using nonlinear gamma correction of the
reflectance component. Four metrics are used to evaluate and compare the experimental results of
the proposed method when implemented on the MEF dataset. The results indicate good
performance of the proposed method.

During the Victorian era many women begin to voice dissatisfaction with their lack of
opportunities and not only try but decide confront and change the bad situation. Other
women rebel by retreating into misery, depression and sometimes madness. It is clearly that
the period after WW1 was the age of the suffragettes that caused the emergence of many
feminist movements; one of them was suffragette movement which created the idea of the
new an independent woman.
America particularly witnesses a great change at the beginning of 20th century. Despite
the WW1 was ended, still its influence affects society as people feels restless due to war’s
stress; looking for a release. New modernity that less restrictive are replaced the old
Victorian sever principles. The people begin to adopt new beliefs, morals and attitudes.
Women have new position in society, wider variety of jobs, and more freedom and equal
rights.
Many dramatists shed light on the suffering of woman and the obstacles that hinder her
from achieving her personality and freeing herself from the prison of Victorian traditional
convictions. Suzan Glaspell (1876–1948) was one of them, an American Pulitzer Prizewinning playwright, novelist, journalist and actress who play a significant role in 20th
century American literature in general. Her plays concentrate on revealing how Victorian
society obliged some women to be imprisoned in roles that might be unfitted for them as
wives and mothers.

The swift progress of artificial intelligence (AI) technologies has
significantly impacted numerous industries, including education. This paper
investigates the use of AI in e-learning management systems (LMS) and its effects on
teaching strategies and outcomes. It explores various AI tools, such as adaptive learning
algorithms, natural language processing, and data analytics, and their role in
personalizing education, streamlining administrative functions, and improving student
engagement and retention. A mixed-methods approach is employed, using both
quantitative data from AI-enhanced LMS implementations and qualitative insights
from educators and students. The findings reveal that AI improves the efficiency of
managing learning processes and tailoring educational experiences, resulting in a more
dynamic and effective learning environment. However, concerns such as data privacy,
the digital divide, and high initial costs are also considered. The paper concludes with
suggestions for AI integration in e-learning platforms, focusing on equitable access and
ethical considerations. The present study plays a role in the broader conversation on the
educational innovation, as it offers a guide for AI utilization for the enhancement of elearning management.

Non-renewable energy is about to be depleted, so renewable energy has emerged as an alternative
that has helped researchers and those interested to snore all the possibilities to make the most of
the factors and natural resources available in the service of humanity. Renewable energy is used
in all military, economic, industrial, agricultural, and all humanitarian activities that serve people
and reduce the use of non-renewable energy. This paper represents a global analysis by regions of
an interest to societies and gives accurate indicators to get rid of dependence on non-renewable
energy sources and protect them from economic fluctuations while pushing sustainable
development in the right direction to create jobs and alleviate poverty. The results of the paper
indicated that there is a significant development in the capacity and production of renewable
energy, and this indicates the presence of interest and follow-up from the countries of the world in
supporting alternative energy, the highest regions were Asia, Europe, the Americas, Oceania,
Middle East, and Africa. The countries with the highest renewable energy capacity and production
in the world were China, USA, Brazil, and Germany. In the field of alternative energy contribution
to electric power capacity and production, the percentages of South American countries were the
highest, followed by Europe, and the lowest were the Middle Eastern countries

This paper tackles sympathetic ideology as a discursive practice in anti-racism speeches
of New Zealand’s Prime Minister, Jacinda Ardern. The paper intends to explore
Ardern’s powerful speeches that convey sympathetic ideologies and societal norms.
The data involves the analysis of two speeches delivered by Jacinda Ardern following
the Christchurch mosque shootings in New Zealand. The study is carried out through
the qualitative research analysis approach relying on (Fairclough, 1995) threedimensional model and (Hyland, 2015) metadiscourse at the interactive and
interactional levels. The paper concluded that the Prime Minister uses certain linguistic
devices that feature her political discourse as in using repetition, hedges and boosters,
which are intended to show solidarity and emphasize social values such as solidarity,
global unity and rejection of racism. Ardern uses the inclusive pronouns “we” and “our”
regularly in her speeches as an engagement marker that reflects a great commitment
towards the population in New Zealand regardless of their racial and cultural
backgrounds.

The increasing use of online services and products in government and business has
led to a big increase in how much people use the Internet, which is managed by
network operating systems. However, this also makes systems more vulnerable to
harmful actions. Recently, there has been a big increase in the need for using
strategies to improve the way we handle attack data to strengthen cybersecurity.
These disciplines include vulnerability assessment, malware categorization,
detection of intrusions, spam detection, and spoofed identifying. This article presents
the crucial techniques for enhancing the selection and extraction of network attack
data, with a specific emphasis on contemporary hybrid methods that use machine and
deep learning algorithms.

The study aims to explore the influence of double letters on handwriting in English compared to Italian. A methodological approach was applied, involving 20 English-speaking participants from Harvard University and 20 Italian-speaking participants from the University of Grenoble. The experiment utilized the Ductus software to analyze handwriting data, including inter-letter intervals and individual letter durations. The study concluded that words with double letters are written faster than other words, with significant differences between the two languages. The effect of double letters was more pronounced in Italian, highlighting the role of linguistic features in influencing motor processes during writing.

Sports language is considered one of the most complex and dynamic areas
within translation studies, as it encompasses unique terminology, culturally
specific activities, and distinct linguistic elements that vary across the two
languages: English and Arabic. This study investigates the translation of English
sports texts into Arabic, focusing on the challenges and variances that arise in
rendering such texts accurately for Arabic-speaking audiences. Five sports texts,
each signifying a different sport, were selected, and ten graduate-level
translators were tasked with translating these texts into Arabic. The researcher
developed an eclectic model grounded in Variational Translation Theory (VTT)
to systematically analyze the translated texts in the target language. The study
hypothesizes that many graduate translators are likely to misuse VTT tools in
translating English sports texts into Arabic. The analysis results support this
hypothesis, revealing that a significant proportion of the translators scuffled to
apply VTT tools accurately, underscoring the challenges in attaining effective
and culturally nuanced translations in the field of sports.

The expansion of the social media environment has created its own linguistic
realities which involve more colloquial communication and practical employment of
language. This research focuses on nominalizations in detail, which are originally formed
words that have been changed for a noun role. These nominalizations are examined within
the context of Facebook posts. The research aims to discover the various nominalizations
used and how often they appear in a large sample of data from Public Facebook Posts
Corpora.
Computational linguistics opened new fields of study and enabled researchers to
study large amounts of data easily, making it easier to identify patterns. Two computational
methods of identifying nominalization in a large dataset were tested and evaluated from
different aspects. Additionally, machine learning analysis tools were conducted to
determine how nominalization affects the complexity and readability of the texts.
Accordingly, Flesch-Kincaid Grade Level and Flesch Readability results were studied to
identify nominalizations’ effect on the texts.
The results present a high density of verbal nominalizations indicating that there is
a preference to refer to actions or processes in an abbreviated manner. Instead, the study
also explores the semantic and functional properties of nominalizations, discovering their
employment for emotions and opinions, focusing on the ideas and informal nature of
Facebook. Despite what the readability formulas imply about the relation between
nominalization density and text difficulty, the research does not discount the fact that these
metrics have their shortcomings, especially when comprehending direct user
understanding is not taken under consideration.
In conclusion, this study adds to the current literature on nominalizations by
identifying their role in the social media discourse and importing important knowledge on
the progressive changes in online linguistic usage.

In this study, boehmite nanoparticles were prepared using aluminum waste extract and their
effect on the photostability of polycarbonate films with a thickness of (5±50) microns after
adding different concentrations of the nanoadjuvant (0.0025, 0.005, 0.01, 0.02, 0.04) % to a
polycarbonate solution dissolved in chloroform. The boehmite nanoparticles were
characterized using various techniques such as EDX, XRD, SEM, FT-IR and TEM. A band at
3200-3500 cm-1 belonging to the OH group was identified according to FT-IR analysis, indicating the formation of boehmite nanoparticles. The photodegradation of polycarbonate
with and without the additive was monitored using UV-visible analysis at a wavelength of
356 nm during different irradiation times (160,120,80,40,20,10,0) hours and the rate constant
of photocatalytic fragmentation of polycarbonate was calculated. The results showed a
decrease in the photodegradation of the polymer films with an increase in the nanomaterial
concentration compared to the films without the additive, Using FT-IR analysis, the carbonyl
absorption coefficient was calculated and found to decrease with the concentration of added
nanobohmite, consistent with the Kd values (photodissociation rate constant). In addition, the
average viscous molecular weight of the polymer, the photodissociation degree, and the
numerical rate of cutting (S) of the polymer chain were calculated using a viscometer. It was
observed that the average molecular weight of the polymer decreases with increasing
irradiation time to a lesser extent in the presence of the nanomaterial additive

This analysis explores how Anne of Green Gables by L.M. Montgomery and F. Scott Fitzgerald’s This Side of Paradise fit the Bildungsroman genre, focusing on the protagonists' journeys from youth to maturity. The objectives include examining the themes of personal growth, identity formation, societal expectations and the development of leadership qualities in both novels. The key research question is: How do Anne and Amory's experiences reflect the conventions of the Bildungsroman, particularly in their development of leadership characteristics, and how do their paths differ based on their environments and challenges? the analysis examines how both characters navigate self-discovery, leadership growth, and the pressures of adulthood by using literary theory on the Bildungsroman. The study concludes that while both protagonists mature and develop leadership traits, their growth is shaped by contrasting cultural and personal factors. It shows the versatility of the genre and the role of leadership development according to Bildungsroman theory.

Object extraction is imperative as it permits the separation and identification of precise objects within
an image, empowering more accurate analysis, recognition, and decision-making in numerous
applications like robotics, surveillance, and medical imaging. In this study, a proposed comparison
study has been introduced between two object-extraction methods, active contour and dilation of
morphology operation based on RGB or YCbCr color spaces components when processed by a Wiener
filter to reduce noise. Then, some measures are used to evaluate the performance of each proposed
technique.

Intertextuality, the relationship between texts, offers a profound lens for understanding the
interconnection of themes, structures and linguistic styles. This study examines the concept of
intertextuality within selected Surahs from the Holy Qur’an, focusing on how their reference
and resonation appear with other Qur’anic verses. By employing a critical discourse analysis,
namely Fairclough’s model of analysis, this research tries to highlight the thematic and
stylistic interconnection that enhance the Qur’anic cohesive narrative and its universal
message. It has been concluded that the selected Surahs are intertextual in so many points.
First, narrating stories is used on purpose and each story of prophets has a relation to other
stories and the aim is believing in Allah the Almighty and worshipping Him. Second, the
themes are portrayed via lexical choices to draw a clear vision to the reader and listener.
Third, the structure of each of the selected Surahs is the same in the sense that they narrate
consecutive stories, which display intertextuality. The findings aim to demonstrate the
Qur’an’s role in fostering dialogue and continuity among Prophets’ faith.

This thesis proposes a human action recognition (HAR) system utilizing Inception V3, a pretrained convolutional neural network (CNN), combined with transfer learning to automate
surveillance tasks. The system aims to mitigate challenges such as human operator fatigue,
high monitoring costs, and limited scalability. Fine-tuned on the UCF-11 dataset, which
includes 11 action categories, the model achieves a notable accuracy of 98%, demonstrating
strong performance in terms of computational efficiency and resilience to variations in lighting
and camera angles. However, the approach primarily focuses on spatial feature extraction and
does not account for temporal dynamics, such as motion and sequence patterns, which are
crucial for understanding complex actions. Additionally, the reliance on a relatively small
dataset limits the system’s generalizability to real-world applications. Despite these limitations,
the model shows promise for deployment in domains like surveillance, healthcare, and robotics.
Future enhancements will involve integrating temporal analysis methods, such as optical flow
or 3D CNNs, and extending validation to larger, more diverse datasets like Kinetics to improve
scalability and real-world applicability.

The recent progress of deep learning has brought us deepfake technology, which can create
highly realistic synthetic content over various modalities, such as fake audio. It's a significant
threat to safety, trust and to the integrity of media particularly when bad actors are creating fake
audio impersonating people. The main research question in which the study is interested is
how to effectively detect deepfake audio using efficient machine learning models that are
reliable.
This study presents a detection approach using Mel Frequency Cepstral Coefficients (MFCC)
as feature extraction to three deep neural network methods: 1D CNN, 2D CNN and a
combination between CNN and LSTM. All the models are trained and tested with a set of the
benchmark datasets to verify the ability of distinguishing the real voice and deep fake samples.
Experimental results demonstrate that all models achieved high detection accuracy, with the
CNN-LSTM model showing the best overall performance due to its ability to capture both
spatial and temporal features. The major contributions of this study are the proposed efficient
deepfake audio detection pipeline, the comparison between deep learning models evaluation,
and the empirical results in favor of MFCC features for achieving better accuracy of detection.

The Travelling Salesman Problem (TSP) is a Classic nondeterministic problem seeks to identify the shortest path which passes through every city in a single trip before arriving back at the starting location. Ant Colony Optimization (ACO) technique provides good results for TSP, but requires significant computational time. In this paper, some steps are presented to resolve the TSP problem. The description of the new suggested method is the partitioning of a problem into a number of subproblems that can be dealt with separately and combinedly solves them for more than once using ACO for each cluster. In this way, the amount of time required for computation is reduced. The experimental results point that the suggested algorithm provides significantly better results than ACO in the aspects of time required for computation and length of the determined optimal path.

Blockchain technologies are responsible for many of the modern world’s unique
experiences, including cryptocurrencies and managing the world’s food supply. Because these
systems are exponentially expanding, as is the data stored by the relevant Blockchains that drive
them. As such, research into this field is also booming and responsible for several key
developments related to Blockchain methodologies. The following paper is an outline of the
topics, challenges and current uses of Blockchain Data Analytics, including predicting the values
of cryptocurrencies (here: Bitcoin) and detecting crimes committed on the Internet.