4th International Conference on Big Data, IOT & NLP (BINLP 2024)

July 20 ~ 21, 2024, Toronto, Canada

Accepted Papers


Stcanetviz: a Visual Analytics Framework for Optimizing an Ocean Currents Prediction Model Stcanet
ABSTRACT

STCANet is a spatio-temporal coupled attention deep network model for predicting ocean currents, which consists of a spatio- temporal module and an attention module , forming a complex nested/coupled system. Understanding and optimizing the STCANet model is challenging due to the complexity of the underlying modules influencing model predictions.In this work we present STCANetViz, a web- based visual analysis tool for understanding and optimizing STCANet models. We proposed an Inter-Module Interdependencies Computing Algorithm (IMDCA), which computes module specific interdependency metrics derived from module interactions to optimize the STCANet model. A modular inter-dependency graph is then designed for examining and analyzing the behavior of the model.The graph consists of a network layer visualization (for module-level exploration) and a heatmap-scatterplot (for module output analysis). We implement a scalable visualization technique incorporating cloud storage for dynamically extracting, seamlessly visualizing and analyzing the high-dimensional ensemble data generated by the STCANet model on ensemble force-directed diagrams (EFDs).

KEYWORDS

Nested / Coupled models, Scalability, Ensemble data, Visual Analytics.


Novel Rotor Fault Diagnostic Method Based on Rlmd and HT Techniques

Asma Guedidi1 and Widad laala2, 1Department of electrical Engineering, Mohamed khider biskra University, Biskra, Algeria, 2Department of electrical Engineering, Mohamed khider biskra University, Biskra, Laboratory of electrical engineering Biskra (LGEB)

ABSTRACT

Frequency-domain analysis using the fast Fourier transform (FFT) has been a popular method for diagnosing broken rotor bar (BRB) faults in squirrel-cage induction motors (IM). However, FFT analysis is limited by sampling frequency and time acquisition constraints, making it less effective under time-varying conditions. This paper introduces a novel BRB fault detection method for nonstationary conditions, building on the recently developed Robust Local Mean Decomposition (RLMD) and Hilbert Transform (HT) techniques. Our proposed method demonstrates that it can accurately track the frequency and amplitude of the 2sf component (where s and f represent the fundamental stator current frequency and motor slip, respectively) under non-stationary conditions. The efficacy of this method is validated through simulations conducted in the Matlab environment. .

KEYWORDS

RLMD, Hilbert transform, fault diagnosis, induction motor.


Machine Learning Innovations in Supply Chain Management: Revolutionizing Predictive Modeling for Efficiency and Growth

Sarthak Pattnaik1, Natasya Liew1, Ali Ozcan Kures1, Kathleen Park2, and Eugene Pinsky1, 1Department of Computer Science, Metropolitan College, Boston University 1010 Commonwealth Avenue, Boston, MA 02215, 2Department of Administrative Sciences, Metropolitan College, Boston University 1010 Commonwealth Avenue, Boston, MA 02215, United States of America

ABSTRACT

Supply Chain Management (SCM) stands at the forefront of modern business practices, serving as the backbone of efficient operations in a globalized and interconnected world. This paper delves into the intricate web of activities that define SCM and explores the transformative impact of technology on optimizing the flow of goods, services, information, and finances across organizations. The evolution of SCM from a logistics-centric function to a multi-functional corporate endeavor underscores the complexity and criticality of managing the end-to-end supply chain processes. By integrating technology into SCM practices, organizations can streamline operations, enhance collaboration, and drive competitive advantage in today’s dynamic marketplace. Technology plays a pivotal role in reshaping traditional supply chain paradigms, offering innovative solutions to longstanding challenges and unlocking new opportunities for growth and efficiency. From advanced data analytics to artificial intelligence (AI) tools, organizations can harness the power of technology to gain real-time insights, forecast demand accurately, and optimize inventory levels. The advent of digital supply chain platforms has revolutionized information sharing and external collaboration, enabling organizations to adapt swiftly to changing market dynamics and customer preferences. Furthermore, technology facilitates the optimization of logistics and transportation processes, leading to reduced fuel consumption, lower carbon emissions, and enhanced sustainability in supply chain operations.

KEYWORDS

Supply Chain Management, Customer Satisfaction, Customer Loyalty, Innovation, Machine Learning, Predictive Models.


Proximal Policy Optimization for Efficient Product Transportation: A Reinforcement Learning Approach

Asharful Islam and Chuan Li, Department of Computer Science, Sichuan University, Chengdu, China

ABSTRACT

Optimizing the delivery of products from a central depot to multiple retail locations presents a multifaceted challenge, especially when considering factors such as minimizing costs while ensuring product availability for customers. Traditional approaches to this problem often rely on heuristic methods or mathematical optimization techniques. However, these approaches may struggle to adapt to dynamic real-world scenarios with complex, evolving conditions. This study pioneers the application of Proximal Policy Optimization (PPO), a state-of-the-art reinforcement learning algorithm, to the domain of product transportation and inventory management. By creating a custom simulation environment, “ProductTransportEnv,” we delve into the complexities of supply chain logistics, demonstrating the significant potential of reinforcement learning to transform operational efficiencies. The “ProductTransportEnv” mimics real-world logistics scenarios, allowing for a detailed exploration of transportation routes, inventory levels, and demand fluctuations, providing a rigorous testing ground for the PPO algorithm.

KEYWORDS

Openai Gym-environment, “ProductTransportEve,” PPO, RL, DQN, Inventory management.


How does e-GP system improve overall quality of e-governance? Based on a conceptual model

Qazi Mahdia Ghyas1, Fumiyo N. Kondo2, 1Dohatec New Media, Dhaka, Bangladesh, 2University of Tsukuba, Japan

ABSTRACT

Every technology should be geared to improve the quality of users lives. Our study aims to understand how electronic government procurement (e-GP) system contributes to quality of electronic governance (e-governance). The current study explored an important issue related to e-GP system improve quality of e-Governance in the Bangladeshi context in order to develop a conceptual framework. Previous research on e-GP adoption has paid attention to manual procurement services. This paper attempts to provide a conceptual model with relationship between two variables: the contribution of e-GP in fifteen specific procurement domains and the contribution of e-GP to overall quality of e-governance. We proposed bottom-up spillover theory with fifteen public procurement domains derived from previous research include: Transparency, Accountability, Corruption Control, Efficiency, effectiveness and predictability, Easy Access, Rules of law and equality, Civil Society Awareness, Fair competition, Online-monitoring , ICT Infrastructure, Employment, Economic development, Social development, Environmental development, COVID-19. The proposed model will contribute to academic literature and provide practical implications, advancing the understanding of e-GP system.

KEYWORDS

e-GP, Quality, e-Governance, bottom-up spillover theory, Bangladesh.


Identification of Bioactive Molecules Present in the Essential Oil of Melicope Madagascariensis (Rutaceae), an Endemic Plant of Madagascar, and Its Biological Activities

Razafindrakoto Fanoina Ny Riana1, 2, 1doctoral School of Process Engineering for Agricultural and Food Industrial Systems, University of Antananarivo, 2antananarivo Higher Polytechnic School, University of Antananarivo

ABSTRACT

The Big Island is one of the few countries in the world that boasts a specific biological megadiversity in terms of flora and fauna. With its exceptional biological diversity, we were guided to focus our study on the leaves of Melicope madagascariensis, belonging to the Rutaceae family, an endemic species, to discover its medicinal properties. The Rue family (Rutaceae) includes more than 125 species across 09 genera in Madagascar, most of which are woody shrubs and trees. The flowers are generally showy and fragrant, and many species have attractive aromatic foliage. Extensive research into the biological activities of the essential oil from the leaves of this plant has been carried out in the rainforest of Madagascars east coast. The essential oil of Melicope madagascariensis has a very specific chemotype. Various GPC and TLC analyses were carried out. The majority of essential oils contain (Z)-β- ocimene (12.7%) and (E)-β-ocimene (25.7%) belonging to the monoterpene family. The synergy of these various bioactive molecules, namely α-pinene (8.56 %); methyl chavicol (6.25 %); trans-piperitol (7.42 %); (E)-β-caryophyllene (3.21 %), gives it several biological activities and an aromatic profile that is unusual for Madagascar. The synergy of its bioactive molecules helps to eliminate free radicals and fungal germs. However, analyses of its toxicity have produced some interesting results. Therefore, the notion of the duality of "efficacity and toxicity" needs to be taken into account, as all active substances are potentially toxic. This is why a more in-depth toxicity study was carried out, using various in vitro tests, namely larval activity and acute toxicity, to determine the lethal dose (LD50) performed on Swiss Musculus L mice. Given its biological properties, the use of the essential oil of Melicope madagascariensis leaves offers a vast opportunity for scientific exploration.

KEYWORDS

Melicope madagascariensis, Endemic species, Essential oil, Biological activity, Toxicity.


The Role of Organic Fertilizers in Growing Maize and Improving the Properties of Saline Soil

Raji .A , AL- Awadi (Ph.D) Erciyes University,Turkey

ABSTRACT

The experiment was carried out on the farm of sayeed. Habib Al-Khatib in the city of Numaniyah, 45 km west of the city of Kut, during the 2024 agricultural season. To study 6 types of organic waste: cows, sheep, and poultry, then a mixture of cow and sheep waste, a mixture of cow and poultry waste, and a mixture of poultry and sheep waste, in an amount of 2 t.ha-1 for each of them, in addition to the comparison sample F0, were given the following symbols, respectively, F1,F2,F3, (F1+F2), (F1+F3), (F2+F3), to know the effect of these fertilizers on the growth and production of Maize plants and soil properties. The land was plowed, ground, and mixed, then the waste was added to the soil with the surface layer 2 t.ha-1, then the corn grains were planted and irrigation began until the end of the experiment. Agricultural operations and crop service were carried out according to the recommendations followed in the region. Soil, plant and yield measurements were taken at the end of the growing season and included. The research results showed that the level (F 2 + F3) of organic Fertilizer from sheep waste mixed with poultry waste resulted in a significant increase in plant height, root length, Ear length, leaf area, weight of 500 grains, yield, and biological yield compared to other levels of fertilizer and organic matter in the soil, while the PH and EC of the soil did not give a significant difference.

KEYWORDS

Maize ,Poultry, Fertilizer, Organic waste, Al-Khatib.


Pan-cancer Analysis of Oncogenic Gata-binding Factor 2 (Gata2) Identifying Prognostics Value and Immunological Function

Xia Wu, The Rockefeller University, New York, US

ABSTRACT

This study comprehensively examines GATA2 expression across 17 cancer types to explore its role in cancer prognosis and immune response. GATA2 showed variable expression in different cancer cell lines and was generally overexpressed in malignant tumors, significantly correlating with patient survival and tumor immune infiltration levels. Analysis revealed associations between GATA2, tumor-infiltrating immune cell markers, and immune response scores. Enrichment analysis identified relevant biological terms and pathways linked to GATA2. The findings suggest GATA2s critical involvement in cancer development and its influence on immune cell infiltration, highlighting its potential as a therapeutic target in cancer immunotherapy. This research supports GATA2s importance as a prognostic biomarker and suggests opportunities for novel targeted cancer treatments.

KEYWORDS

pan-cancer, prognostic biomarker, immune infiltration, GATA2, TME.


Biologics and Biosimilars in Albania; Consumption and Costs

Dorela Hasankolli1, Mirela Miraci1, Teuta Backa2, Marsida Mulaj3, 1Departament of Pharmacy, University of Medicine, Tirana, Albania, 2Department of Rheumatology, "Mother Teresa" University Medical Center, Tirana, Albania, 3Compulsory Health Care Insurance Fund, Tirana, Albania

ABSTRACT

Biological therapies use proteins that have been produced with biological technology to produce targeted effects within the body, impact the bodys immune system, and decrease inflammation at different stages.Biologics have been utilized in the treatment of conditions like rheumatoid arthritis, systemic lupus, ankylosing spondylitis, arthropathic psoriasis, and psoriatic arthritis.The aim of the study is to conduct a comprehensive review and analyze the market dynamics related to the utilization of these treatments in Albania. Methods: This is a retrospective and observational study based on FDSKSH’s data for 2023. Results: The total annual reimbursement cost for 2023 was ALL 188,208,780 (1ALL =0.01 USD).The average reimbursement cost per patient was ALL 209820.2679. The use of biologics and biosimilars represents only 0.028% of all prescriptions and accounts for 1.56% of the total reimbursement costs for drugs.In 2023, the reimbursement cost covered by FDSKSH was 7.14% lower than in 2022.87.17% of the patients were treated with Tumor Necrosis Factor (TNF) Inhibitors. Etanercept 25 mg (biosimilar) had the highest price reduction of 32.83% in 2024. The most expensive drug in 2023 was Secukinumab 150 mg. Biologics account for 61% of the total annual reimbursement cost, while biosimilars account for 39%. Conclusions: The prescription of biologics is lower compared to biosimilars,but they are associated with significant costs that impact the overall reimbursement costs.The prices of biologics and biosimilars were lower than those in Kosovo, Italy, Greece and will decrease in 2024. Cost-economic evaluations of the most expensive biologics and biosimilars are important for the management of financial resources.

KEYWORDS

Biological therapies, the total annual reimbursement cost, the average reimbursement cost per patient, prices, biosimilars.


A Budget Impact Analysis of Adalimumab Biosimilars; the Albanian Context

Dorela Hasankolli1, Mirela Miraci1, Teuta Backa2, Entela Haloçi1, 1Departament of Pharmacy, University of Medicine, Tirana, Albania, 2Department of Rheumatology, "Mother Teresa" University Medical Center, Tirana, Albania

ABSTRACT

Adalimumab is used to treat various autoimmune conditions such as rheumatoid arthritis, ankylosing spondylitis, psoriasis, psoriatic arthritis, Crohn’s disease, and ulcerative colitis.Adalimumab isa tumor necrosis factor (TNF) inhibitor. Biosimilars of adalimumab have the potential to reduce costs for the healthcare system. The aim of the study is to develop a drug budget impact analysis to assess the potential financial impact of the availability of thebiosimilar adalimumab-aacfin terms of cost savings due to its lower price compared to adalimumab-adaz. Methods: This is a budget impact analysis and cost calculator model in Microsoft Excel based on FDSKSH’s data for 2023. Results:Adalimumab biosimilars accounted for approximately 25% of biologic drug prescriptions in 2023. The reimbursement cost of biologic drugs for autoimmune diseases amounted toALL 186,523,409 in 2023. Of that cost, 15.21% accounted for treatment with adalimumab biosimilars. Adalimumab-aacf was available at a 15.82% lower price than adalimumab–adaz. Analyses indicated that the replacement of the biosimilar adalimumab-adaz with the biosimilar adalimumab-aacfcould result in cost savings of ALL 968,138, and thereimbursement cost of biologic drugs for autoimmune diseases could have been 3% lower. With the potential cost savings, two more patients could have had access to biologic drugs in 2023.Conclusions: The Drug Budget Impact Analysis estimated that the availability of adalimumab-aacf would be more profitable in terms of cost savings compared to adalimumab-adaz. The scenario could result in increased treatment access for patients.

KEYWORDS


Evaluation of Galaxy as a User-friendly Bioinformatics Tool for Enhancing Clinical Diagnostics in Genetics Laboratories

Hadi Almohab1 and Ramzy Al-Othmany2, 1Faculty of Engineering, Computer, and Design, Nusa Putra University, Sukabumi 43152, Indonesia, 2Department of Biochemistry, Institute Pertanian Bogor University. Bogor ,Indonesia

ABSTRACT

Bioinformatics platforms have revolutionized clinical diagnostics by facilitating the analysis of genomic data, thereby advancing personalized medicine and enhancing patient care. This study investigates the integration, usage patterns, challenges, and impact of the Galaxy platform within clinical diagnostics laboratories. Employing a convergent parallel mixed-methods design, quantitative survey data and qualitative insights from structured interviews were gathered from fifteen participants across diverse roles in clinical settings. Findings reveal widespread adoption of Galaxy, with high satisfaction reported for its user-friendly interface and significant improvements in workflow efficiency and diagnostic accuracy. Challenges such as data security and training needs were identified, emphasizing the platforms role in simplifying complex data analysis tasks. The study contributes to understanding Galaxy’s transformative potential in clinical practice, offering recommendations for optimizing its integration and functionality. These insights are crucial for advancing clinical diagnostics and enhancing patient outcomes.

KEYWORDS

Bioinformatics, Galaxy, Diagnostics, Precision, Genomics, Workflow, Accuracy, Cybersecurity.


Quantum Reinforcement Learning for High Frequency Trading

Alexander Kirnasov, Head of Quant Dev in Znamenka Capital

ABSTRACT

We introduce a new approach of Reinforcement Learning Application for High Frequency Trading called Quantum Reinforcement Learning as our agent learns to react on ‘quantum’ individual events in Limit Order Book – single Limit Order Book updates and single trades (and optionally single Orders if provided by Exchange). We claim that such level of learning granularity allows our agent to find optimal trading strategies by on-line modeling of Market Microstructure with a maximum rate and precision.


Predicting Drug-drug Interactions using Knowledge Graphs

Lizzy Farrugia, Lilian M. Azzopardi, Jeremy Debattista, and Charlie Abela, Faculty of ICT & Faculty of Medicine and Surgery, University of Malta, Malta

ABSTRACT

In the last decades, people have been consuming and combining more drugs than be- fore, increasing the number of Drug-Drug Interactions (DDIs). To predict unknown DDIs, recently, studies started incorporating Knowledge Graphs (KGs) since they are able to capture the rela- tionships among entities providing better drug representations than using a single drug property. In this paper, we propose an end-to-end framework that integrates several drug features from public drug repositories into a KG and embeds the nodes in the graph using various translation, factorisation and Neural Network (NN) based KG Embedding (KGE) methods. Ultimately, we use a Machine Learning (ML) algorithm that predicts unknown DDIs. Among the different translation and factorisation-based KGE models, we found that the best performing combination was the ComplEx embedding method with a Long Short-Term Memory (LSTM) network, which obtained an F 1-score of 95.19% on a dataset based on the DDIs found in DrugBank version 5.1.8. This score is 5.61% better than the state-of-the-art model DeepDDI. Additionally, we also developed a graph auto-encoder model that uses a Graph Neural Network (GNN), which achieved an F 1-score of 91.94%. Consequently, GNNs have demonstrated a stronger ability to mine the underlying se- mantics of the KG than the ComplEx model, and thus using higher dimension embeddings within the GNN can lead to state-of-the-art performance.


Learning to Play Atari Games Using Dueling Q-learning and Hebbian Plasticity

Ashfaq Salehin, University of Sussex, UK

ABSTRACT

In this work, an advanced deep reinforcement learning architecture is used to train neural network agents playing atari games. Given only the raw game pixels, action space, and reward information, the system can train agents to play any Atari game. At first, this system uses advanced techniques like deep Q-networks and dueling Q-networks to train efficient agents, the same techniques used by DeepMind to train agents that beat human players in Atari games. As an extension, plastic neural networks are used as agents, and their feasibility is analyzed in this scenario. The plasticity implementation was based on backpropagation and the Hebbian update rule. Plastic neural networks have excellent features like lifelong learning after the initial training, which makes them highly suitable in adaptive learning environments. As a new analysis of plasticity in this context, this work might provide valuable insights and direction for future works.


Sentiment Analysis for Ethiopia’s Tigray War From Twitter Tweets After Pretoria Deal: Measurement of Hope and Fear

Zebene Hailye1, Tigist Mintesnot2, Yelkal Mulualem2, Melaku Bitew2, Abebech Jenber2 and Azanu Mirollign2, 1College of Informatics, department of Information Science, University of Gondar, Gondar, Ethiopia, 2College of Informatics, department of Information Technology, University of Gondar, Gondar, Ethiopia

ABSTRACT

The Ethiopian Tigray War is considered the deadliest conflict of the 21st Century, which devastated Tigray, and ended when the FDRE and TPLF signed a permanent cessation of hostilities in Pretoria on 3 November 2022. This research presents a new lexicon-based unsupervised sentiment analysis technique to assess hope and fear of the 2021 Tigray war in Ethiopia. The top 54 tweets and related comments from two distinct Twitter (X) feeds about Ethiopias Tigray war news (Getachew K Reda and Redwan Hussien) are scraped and a data set is built. We advocate to apply a dictionary technique that rates the optimism of each submitted user account. Additionally, a topic modeling technique, Latent Dirichlet Allocation (LDA) algorithm applied to comprehend the primary concerns brought up by users and the key discussion points. Experiment results show that hope increases significantly following the Pretoria deal, however, theres also slight increasing of fear.

KEYWORDS

Hope, Fear, Sentiment Analysis, Topic Modeling, Ethiopia, Tigray, War.


Optimizing Ground Sampling Distance for Drone-based Gis Mapping: a Case Study in Riyadh, Saudi Arabia

Moghaid Farah, and Abdulaziz Alruwaili, WAKAB Company, Department of Unmanned Aerial Vehicles, Alyasin, Riyadh 13521, Saudi Arabia

ABSTRACT

Drone-based Geographic Information Systems (GIS) mapping has gained significant popularity due to its cost-effectiveness and efficiency in capturing high-resolution aerial imagery. The Ground Sampling Distance (GSD) plays a crucial role in the accuracy and quality of dronebased mapping. This paper explores the relationship between GSD and GIS mapping accuracy using drone imagery. It investigates the trade-off between GSD, flight altitude, and image resolution to determine the optimal GSD for different mapping applications, through a case study in Riyadh, Saudi Arabia. Furthermore, to identify the optimal altitude for a drones camera in different applications, a new factor is introduced. This factor aims to determine the most suitable flight altitude that achieves the highest level of accuracy and quality in the captured imagery.


Analyzing the Impact of Increased Spatial Mobility on Sustainability: Solutions and Personal Reflections

Issac Hsu, Aalto University, Finland

ABSTRACT

This paper examines the sustainability challenges posed by the increased movement of people and objects, driven by advancements in transportation technologies. Through a comprehensive review of literature and mobility data analysis, we explore the environmental impacts of heightened transportation activities, including greenhouse gas emissions, air and noise pollution, and habitat disruption. The study identifies multifaceted solutions to address these issues at urban, interurban, and global scales, emphasizing the role of electric and hydrogen-fuel-cell vehicles, public transportation systems, and smart traffic management. Additionally, the paper reflects on the authors travel behaviors, highlighting the significance of individual choices in promoting sustainability. The challenges and methodologies for analyzing mobility data from GNSS devices, mobile phone operators, and social media platforms are discussed, providing insights into the complexities of data privacy, accuracy, and representativeness.

KEYWORDS

Spatial Mobility, Sustainability, Greenhouse Gas Emissions, Urban Transportation, Data Analysis, Personal Travel Behavior.


IOT-based Yield Prediction System for Better Vegetable Production

Asia Mumtaz1, Ahmed Shukri Mohd Noor1, Saba Farzeen23, 1Faculty of Ocean Engineering Technology and Informatics, University Malaysia Terengganu, Kuala Terengganu, 21300, Malaysia, 2Department of Computer Science, University of Sialkot, Sialkot, 51310, Pakistan, 3Department of Computer Science, University of Sialkot, Sialkot, 51310, Pakistan

ABSTRACT

In modern agriculture, accurate crop yield prediction is crucial for optimizing food quality, production, and profitability while minimizing costs. This research focuses on developing a yield prediction system for enhancing vegetable production, by leveraging IoT-based sensors. The study collects comprehensive soil data, including temperature, humidity, soil moisture, light intensity, pH, and internet-sourced weather data. Machine learning algorithms are employed to analyze this dataset and accurately forecast crop yield. The systems precise and timely predictions empower farmers to make informed decisions on irrigation, fertilizer WebApp, and pest control, ultimately leading to improved vegetable production. Implementing IoT-based sensors in Sialkot showcased significant improvements, resulting in higher profitability for farmers. With higher accuracy achieved through the Random Forest model, compared to the Support Vector Machine and Decision Trees, this research provides foundational basis for developing yield prediction systems in agriculture, offering solutions for global farmers to enhance crop management practices sustainably and efficiently.

KEYWORDS

ML, IoT, Random Forest, SVM, Decision tree.


Securing the Future of Healthcare: Building a Resilient Defense System for Patient Data Protection

Ejiofor Oluomachi, Akinsola Ahmed, Department of Computer Science, Austin Peay State University, Clarksville USA

ABSTRACT

Data is becoming increasingly important within the healthcare sector, which has led to increased cybercrime directed towards the stealing of vital information of patients. Loss of data has cost many healthcare organizations, particularly clinics and hospitals. This investigation aims to propose the ideal approach to developing a defense system that ensures that patient data is protected. More specifically, the study aimed to improve cyber health through prevention of intruders. The study adopted the machine learning model of gradient boosting classifier to predict the severity of breaches of healthcare data. Secondary data was collected from U.S. Department of Health and Human Services Portal with key indicators. Also, the study gathers key cyber-security data from Kaggle, which was utilized for the study. The findings revealed that hacking and IT incidents are the most common type of breaches in the healthcare industry, with network servers being targeted in most cases. The model evaluation showed that the gradient boosting algorithm performs well. Therefore, the study recommends that organizations implement comprehensive security protocols, particularly focusing on robust network security to protect servers.

KEYWORDS

Cyber Health, Cyber Security, Defense System, Patient Data Protection. Healthcare Data.


Sustainable Investments and Esg: Portfolio Optimization Using Genetic Algorithms

Larissa Luize de Faria Cardoso, Electrical Engineering Department Pontifical Catholic University of Rio de Janeiro, Rio de Janeiro, Brazil

ABSTRACT

Sustainable investments, guided by ESG (Environmental, Social, and Governance) criteria, have become a central focus for investors worldwide. The integration of ESG criteria into investment decisions has been shown to lead to better financial performance and lower long-term risk for companies. This study aims to develop and apply a Genetic Algorithm (GA) to optimize investment portfolios that balance financial, return, ESG criteria, and risk. The proposed methodology creates a robust and adaptable model suitable for real-world sustainable investment scenarios. By using data from companies such as Apple, Microsoft, and Tesla, this study demonstrates the effectiveness of GAs in achieving an optimal portfolio allocation. The results highlight the potential of GAs to consider multiple objectives simultaneously and provide a balanced solution that meets financial and sustainability goals.

KEYWORDS

Sustainable Investments, ESG, Genetic Algorithms, Portfolio Optimization, Financial Performance.


A Survey on Prompt-free Few-shot Text Classification Performance and Limitations

Rim Messaoudi, Achraf Louiza Rim and Francois Azelart, Akkodis Research - Akkodis France

ABSTRACT

Text-based comments play a crucial role in providing feedback for various industries. However, effectively filtering and categorizing this feedback based on custom context-specific criteria requires sophisticated language modeling techniques. While traditional approaches have shown effectiveness, they often require a substantial amount of data to compensate for their modeling deficiencies. In this work, we focus on highlighting the performance and limitations of prompt-free few-shot text classification using open-source pre-trained sentence transformers. On the one hand, our research includes a comprehensive study across different benchmark datasets, encompassing 9 dimensions such as sentiment analysis, topic modeling, grammatical acceptance, and emotion classification. On the other hand, we underline prompt-free few-shot classification limitations when the targeted criteria are complex. As an alternative approach, prompting an instruction-fine-tuned language model has demonstrated favorable outcomes, as proven by our application in the specific use case of “Identifying and extracting resolution results and actions from explanatory notes”, achieving an accuracy rate of 80%.

KEYWORDS

Language models, Sentence transformers, SetFit, contrastive learning, distillation, intelligence compression, NLP, semantic similarity.


Addressing the Limitations of News Recommendation Systems: Incorporating User Demographics for Enhanced Personalization

Zerihun Olana Asefa and Admas Abtew, Jimma University, Department of Information Technology

ABSTRACT

News recommendation schemes utilize features of the news itself and information about users to suggest and recommend relevant news items to the users towards the interest they have. However, the effectiveness of the existing news recommendation scheme is limited in the occurrence of new user cold start problems. Therefore, we designed a news recommender system using hybrid approaches to address new user cold start problems to ease and suggest more related news articles for new users. To achieve the objective mentioned above, user demographic data with a hybrid recommendation system that contains the scheme of both content-based and collaborative filtering approaches is proposed. To evaluate the effectiveness of the proposed model, an extensive experiment is conducted using a dataset of news articles with user rating value and user demographic data. The performance of the proposed model is done by two ways of experiment. So, the performance of the proposed model performs around 68.05% of Precision, 42.46% of Recall and 52.1% of the average F1 score for the experiment based on individual user similarity in the system. And also performs around 93.75% of precision, 40.25% of recall and 56.31% F1-score for the similarity of users based on the similarity of users within the same category which is better than the first experiment.

KEYWORDS

news recommendation system; cold start problem; hybrid approach; demographic information; new users; popular news.


Mapping Chatgpt in Mainstream Media: Early Quantitative Insights Through Sentiment Analysis and Word Frequency Analysis

Maya Karanouh, Phd Candidate, Department of Interdisciplinary Humanties, Brock University, Canada

ABSTRACT

The exponential growth in user acquisition and popularity of OpenAI’s ChatGPT, an artificial intelligence (AI) powered chatbot, was accompanied by widespread mainstream media coverage. This article presents a quantitative data analysis of the early trends and sentiments revealed by conducting text mining and NLP methods onto a corpus of 10,902 mainstream news headlines related to the subject of ChatGPT and artificial intelligence, from the launch of ChatGPT in November 2022 to March 2023. The findings revealed in sentiment analysis, ChatGPT and artificial intelligence, were perceived more positively than negatively in the mainstream media. In regards to word frequency results, over sixty-five percent of the top frequency words were focused on Big Tech issues and actors while topics such as jobs, diversity, ethics, copyright, gender and women were poorly represented or completely absent and only accounted for six percent of the total corpus. This article is a critical analysis into the power structures and collusions between Big Tech and Big Media in their hegemonic exclusion of the “Other” from mainstream media.

KEYWORDS

ChatGPT, Generative AI,Artificial Intelligence, News headlines, Bias, NLP, Data Analytics, Sentiment analysis.


Beyond Borders: Empowering Multilingual forms with Generative AI using Marianmtmodel and T5 Model

Jayansh sharma1 and Rituparna Datta2, 1Indian Institute of information Technology Una, 2Capgemini Technology, India

ABSTRACT

In a world where connecting and working with people from different countries is more and more important, the language barriers are often the main reasons why the cross border communication and collaboration is not successful. This research paper is about the use of Generative AI models, most notably the MarianMTModel and T5 Model, that enable to go through the linguistic boundaries and create the multilingual forms. The paper, on the other hand, explores the real-life application of these models in a Python environment through the Hugging Face Transformers Library. The paper goes into detailed code sample to show how these models can be used to brightly transfer textual data from one language to another apart from currently utilized models . The experimental design concerns with the translation of different sample data, this data contains individual attributes like name, age, height, weight, and the medical problems, into a number of target languages. Besides, this study not only shows the technical difficulties of model initialization and translation but also it emphasizes the wider meaning of such technology for developing cross-cultural understanding and making the world communication easier. The results underline Generative AIs potentiality to overcome language obstacles, thus enabling the worldwide cooperation, knowledge spread, and cultural exchange.

KEYWORDS

Generative AI, MarianMTModel, T5 Model, Natural Language Processing, Hugging Face Transformers, Cross-cultural Communication, Language Barriers, Computational Linguistics, Multilingual Translation.