Authors - K. N. Subramanya, Padmashree T., Manojith Bhat V., Manasvini G. Padmasali Abstract - In an era of rising digital dependence and ever-evolving cyber threats, cybersecurity has become a concern for education institutions of all sizes. Higher education institutions (HEIs) are prime targets for cyberattacks since they manage massive volumes of sensitive data related to students, faculty, and research. Protecting this data is important to avoid major consequences such as disruptions in academic services, reputational harm, legal or financial ramifications. This survey consolidates current research on cybersecurity practices in HEIs, analyzes critical digital assets and infrastructure. It also reviews selected cybersecurity frameworks adopted globally. The survey further explorers the threat landscape confronting HEIs, examining various cyberattacks by identifying possible entry points, attack pathways, and potential consequences. The study emphasizes the necessity of adaptive cybersecurity approaches that can evolve alongside emerging technologies and pedagogical models in academia.
Authors - Roberts Dargis, Arturs Znotins, Ilze Auzina, Maris Golubovskis, Mikelis Gulbis, Normunds Gruzitis Abstract - Operational radio communication is a challenging application domain for automatic speech recognition (ASR) despite advances in multilingual foundation models. We investigate the applicability of modern Latvian ASR models to operational communication and evaluate whether speech enhancement techniques improve recognition quality under realistic conditions. To support the study, we created a specialized corpus of authentic Latvian operational radio communication. The corpus captures acoustic and linguistic phenomena largely absent from general speech corpora, including narrow-band transmission, radio-channel artifacts, environmental noise, domain-specific terms, and fragmented utterances. Using this corpus, we evaluate state-of-the-art adaptations of the massively multilingual Whisper and MMS models in combination with several audio preprocessing methods. The results reveal a substantial performance gap between the conventional Latvian ASR benchmarks and operational communication data. While some preprocessing methods improve perceived audio quality, they provide limited benefit for downstream recognition and often even degrade ASR performance. Voice activity detection, however, yields the most consistent improvements. The findings indicate that domain mismatch, rather than acoustic degradation alone, is the dominant source of recognition errors and highlight the need for representative domain-specific data when adapting general-purpose ASR models for the operational communication environment.
Authors - Mark Fedorchenko, Olena Kopishynska, Yurii Utkin, Igor Sliusar, Leonid Flehantov, Olha Barabolia, Nadiia Protas, Tetiana Dugar Abstract - Crop yield forecasting based on small official statistics is different from forecasting with dense satellite, field, or weather datasets: the sample is short, temporal leakage is easy to introduce, and machine learning (ML) should not be accepted unless it beats transparent baselines. This paper presents a baseline-first and reliability-aware workflow for farm management and regional advisory systems. Wheat, maize, and sunflower are evaluated for Poltava, Vinnytsia, Cherkasy, and national-level Ukraine data for 2010-2024. ElasticNet, XGBoost, and LightGBM are compared with naive lag-1, linear-trend, LINEST, and Autoregressive Integrated Moving Average (ARIMA) baselines under a forward temporal design. The contribution is a decision layer that recommends ML only after it clears a practical mean absolute error (MAE) margin and then reports empirical validation-residual bands, test coverage, feature-group diagnostics, and compact farm management systems (FMS)-compatible forecast cards. The Poltava workflow recommends FORECAST.LINEAR for wheat (MAE 0.49 t/ha), LightGBM for maize (MAE 0.69 t/ha), and LightGBM for sunflower (MAE 0.04 t/ha). Across the external check, ML is recommended in 7 of 12 region-crop cases. The results show that ML can help in small official-statistics settings only when checked against simple baselines and reported with reliability diagnostics.
Authors - Ingy Emara, Rawan Waleed, Sherry Emad Abstract - This study analyses speech errors in individuals with Down syndrome (DS) in both Arabic and English, with a particular focus on errors that reduce intelligibility for automatic speech recognition (ASR) systems. It compares the speech errors that most significantly affect intelligibility in each language and investigates the factors underlying differences in ASR accuracy across Arabic and English DS speech. The findings indicate that ASR systems perform less accurately with Arabic DS speech, highlighting the need for larger and more diverse Arabic DS speech datasets for system training. The study also identifies several physiologically related speech errors that negatively affect intelligibility in both languages, including devoicing of stop consonants, lateralization of /r/, reduced pressure in /s/, and deletion of consonants and consonant clusters. In addition, certain errors were found to be language specific, such as the mispronunciation of uvular and pharyngeal sounds in Arabic and the frequent omission of /r/ and vowel centralization in English. These findings have important implications for speech therapy by identifying priority areas for intervention, and for the ASR industry by emphasizing the need to expand labelled DS speech datasets across languages to improve recognition accuracy.
Authors - Hiep Nghia Phan, Duy Nguyen Ngoc Abstract - Low-Rank Adaptation (LoRA) is widely used for parameter-efficient fine-tuning of large language models because it is lightweight, modular, and easy to distribute. However, the growing practice of sharing third-party LoRA modules also creates security concerns. A malicious adapter can introduce hidden behaviors, backdoors, or other risks while appearing to function normally. Existing research has largely focused on detecting whether a LoRA module is malicious. In practice, deployment decisions often require a more nuanced assessment of risk. This paper presents a measurable framework that evaluates LoRA security across four dimensions: supply-chain integrity, static weight characteristics, dynamic behavior, and deployment-time observations. The resulting indicators are normalized and combined into a composite risk score that supports comparison and prioritization of LoRA modules. The framework was evaluated using benign and backdoored LoRA modules attached to a frozen base language model. The results show a clear separation between the two groups even when their task performance remains similar. Dynamic behavioral testing and static weight analysis contribute the most useful signals, while deployment-time monitoring provides additional evidence of long-term operational risk. The proposed framework provides a practical mechanism for integrating security assessment into LoRA selection, governance, and deployment workflows.
Authors - Mayen Ben-Koko, Emmanuel Waribo Otiti Abstract - Nigeria loses more new-borns in the first month of life than almost any other country in the world, yet no machine learning tool has been built specifically for this context. This paper proposes a framework for predicting neonatal mortality risk in Nigeria using indicators from the 2023–24 Nigeria Demographic and Health Survey — the most current national health dataset available. Five classification algorithms are compared: Logistic Regression, Decision Tree, Random Forest, Gradient Boosting, and Support Vector Machine. Random Forest performed best, with an AUC-ROC of 0.89. The three strongest predictors were whether a skilled health worker attended the birth, the gap between pregnancies, and the number of antenatal visits. The framework is reproducible and designed to be extended as fuller microdata becomes available or adapted for routine clinic records across Nigeria's six geopolitical zones.
