The extreme volatility of Bitcoin markets makes accurate price prediction notably difficult. This paper proposes a new hybrid deep learning model that incorporates a Gated Recurrent Unit (GRU), a Bidirectional Long Short-Term Memory (Bi LSTM) model, and a Multi Head Attention mechanism to permit the model to utilize both historical price data and sentiment information from Twitter. We constructed the model utilizing a two-stage transfer learning approach: we first pretrained the model on data from 2017−2019 to learn lower-level fluctuation behaviors, then we fine-tuned the model on data from 2021−2023 in order to be sensitive to recent market behaviors. The model performed exceptionally well against multiple state-of-the-art baselines using root mean square error (RMSE) and mean absolute error (MAE) metrics, reporting RMSE values of 679.61 and MAE of 452.95, achieving considerable improvement over the baseline models. Our experimental results show that leveraging Twitter sentiment greatly improved trend prediction. In addition, our benchmarks showed that our method performed better than the existing methods. Furthermore, our ablation studies illustrated how each particular feature performed. Overall, our results demonstrate that multi-scale temporal modeling combined with social media sentiment integration produces a scalable and resilient solution to combat the challenges of volatility to forecast cryptocurrency prices accurately and efficiently.

Many study programs at universities face issues, including students experiencing delays in graduation, which hinders the completion of their studies on time. These delays in student graduation contribute to a decrease in the accreditation score of the Information Systems program. One solution to address this issue is to develop a data-mining-based system to monitor and utilize student progress data by predicting their graduation status using the C4.5 Decision Tree algorithm. This research process involves several stages: problem analysis, data and system design, coding, testing, and finally, maintenance. The outcome of this research is the implementation of the C4.5 algorithm to predict students' timely and delayed graduation. The data used includes records of students who graduated in 2021 and 2022. The acceptance rate, calculated using a confusion matrix, demonstrates an accuracy level of 92.16%, based on a dataset of 119 training data points and 51 testing data points, or 70% training to 30% testing ratio. The results of this research and testing indicate that the C4.5 Decision Tree algorithm is highly suitable for predicting student graduation outcomes.

With the rise of online platforms, concerns are increasing about the presence of fake user profiles, which can be exploited for malicious activities such as fraud, identity theft, and spreading misinformation. This study provides a detailed analysis of four machine learning algorithms to detect fake profiles: Support Vector Machine, Logistic Regression, Passive Aggressive, and Decision Tree. To train and evaluate these models, we first collect a broad dataset of both genuine and fake user profiles. Through feature engineering, relevant data such as text content, account creation details, and behavioral patterns are extracted from the profiles. Support Vector Machine is selected for its capacity to manage high-dimensional data and reduce the risk of overfitting, while Logistic Regression is valued for its interpretability and capability to model complex relationships. Passive Aggressive is included to test performance in real-time scenarios, where fake profile characteristics may evolve due to its adaptability to changing data streams. Decision Trees are employed for their ability to capture non-linear relationships and offer insights into the decision-making process. Metrics like recall, accuracy, and precision are used to evaluate the performance of each algorithm. This comparative analysis enhances our understanding of machine learning approaches for detecting fake profiles and offers practical insights for developers aiming to mitigate risks associated with online fraud. Among the algorithms, Decision Tree achieved the highest accuracy at 98.76%.

The surge in scholarly articles on e-Commerce mirrors its rapid ascent in the market's legitimacy. According to customer product recommendation theory, e-Commerce research may exhibit a bias toward specific customer product recommendations due to its evolving nature. To address this concern, this study examines five of the leading e-Commerce journals. The findings reveal a predominant focus on two main groups: customers and the integration of artificial intelligence (AI) in e-commerce recommendation systems. However, there is a notable lack of attention toward other critical groups, such as suppliers, indirect stakeholders, investors, and regulators. With e-Commerce continuing to mature, it is crucial to explore these neglected themes, sectors, and entities. This paper identifies gaps in current research through targeted keyword searches by aiming to bring these overlooked areas to the forefront. By highlighting persisting challenges in e-Commerce research, this study seeks to raise discourse and innovation in the field by ensuring that emerging topics are not overlooked. The role of AI in e-Commerce, particularly in the development of advanced recommendation systems, is identified as a key area shaping consumer experiences and market dynamics.

Natural disasters cause economic instability, leading to severe financial hardships for affected communities. The rapid surge in essential goods prices during such events significantly burdens vulnerable populations, highlighting the critical need for timely policy interventions. While understanding public sentiment on economic distress is crucial for effective data-driven policy generation, research specifically analyzing public sentiment on price hikes in such contexts remains limited, often due to a lack of dedicated datasets. To address this, this paper first introduces a novel dataset of social media comments on price hikes related to the 2023 Turkey earthquake. Second, to support data-driven policy-making by quantifying public sentiment, we applied a range of AI models and identified transformer-based models like DistilBERT as particularly effective for sentiment classification. Furthermore, we employ Explainable AI techniques to enhance model trust, enabling policymakers to confidently use these insights to support disaster recovery and economic stabilization in affected regions.

Bontang Mangrove Park located in Kutai National Park, North Bontang District, Bontang City, serves as a prominent mangrove forest tourism destination. Despite its popularity, the park lacks an official website, relying solely on social media for information dissemination. This limitation restricts the park's ability to reach a broader audience and provide comprehensive details about its facilities, operational hours, and attractions. To address this issue, a dedicated website was developed to enhance the park's online presence and improve visitor accessibility to information. The website design and development followed the Rational Unified Process (RUP) methodology, an iterative and incremental approach to software development that ensures adaptability to changing functional requirements. Functionality testing was conducted using the Black-box method, while user satisfaction with the website's design and usability was assessed through a Likert-scale questionnaire distributed to 100 participants, indicating a positive reception with 74.03% of respondents rating the website favorably. This rating classifies the website as "Good" in terms of functionality and user experience, demonstrating its potential as a valuable tool for promoting mangrove ecotourism in Bontang Mangrove Park. The findings highlight the website's ability to improve accessibility, promote ecotourism, and engage visitors through digital means.

This study investigates the application of machine learning methods for the classification of fraudulent job postings in e-business platforms. Using the publicly available fake_job_postings.csv dataset, textual and categorical features of vacancies were processed and vectorised through TF-IDF, HashingVectorizer, and optimised TF-IDF. Eight machine learning algorithms were compared, including Logistic Regression, Random Forest, Gradient Boosting, Decision Tree, Multinomial Naive Bayes, Linear SVC, K-Nearest Neighbours, and XGBoost. The experiments demonstrate that XGBoost achieved the best performance (Accuracy = 0.990, Precision = 0.982, Recall = 0.998, F1 = 0.990) across all feature representations. Its superior results can be attributed to the ability of boosted ensembles to capture complex non-linear relationships in high-dimensional feature spaces while maintaining robustness against noise and class imbalance.
However, it should be noted that the evaluation was performed on a single static dataset. While the high recall shows the model’s ability to reliably detect fraudulent ads in this context, questions remain about its generalisability. Fraud tactics evolve rapidly, and new job scams may significantly differ from patterns in the training data. This creates  a potential risk of overfitting to dataset-specific features, which limits direct transfer to real-world scenarios without continuous retraining and monitoring. The practical contribution of the study is a reproducible framework that integrates text and categorical processing, vectorisation, hyperparameter optimisation, and comparative model benchmarking. Such a framework could be embedded into online job platforms to support automated filtering of suspicious ads. Still, its deployment requires additional measures: periodic retraining with updated data, integration with platform APIs, and the inclusion of explainability modules to ensure transparency and user trust. Overall, the research demonstrates that ensemble-based models, particularly XGBoost, offer strong potential for fraud detection in the e-business labour market. At the same time, further work is necessary to validate model robustness on unseen and evolving fraudulent job posting strategies, ensuring scalability and reliability in production environments.