Pronab Pal, Keybyte Systems Intentix Lab, Australia
Modern cloud-native applications distribute business logic across multiple layers: application code, orchestration frameworks, service meshes, and infrastructure configurations. This distribution creates ”hidden logic”—execution rules embedded in infrastructure that are invisible during design and difficult to trace at runtime. We present Intention Space , a computing model built on the CPUX (Common Path of Understanding and Execution) paradigm that consolidates all business logic into explicit, design-time declarations using plain-language state pulses. In our model, Design Nodes (DNs) contain computation while Gatekeepers declare execution conditions as named pulses (e.g., ”payment validated”: Y). The infrastructure provides only mechanical enforcement through an Intention Loop that matches runtime state to Gatekeepers without adding decision logic. We demonstrate that complex workflows—traditionally requiring nested if-then branching and explicit loops—can be expressed as linear CPUX sequences where execution paths emerge from data state rather than code branching. Our Golang implementation shows complete elimination of orchestration code while maintaining full cognitive traceability. Beyond technical innovation, CPUX addresses a critical social computing crisis: the lack of accountability in distributed social platforms. By creating unique, device-level CPUX footprints for every interaction, our model enables verifiable traceability from device identity through user intention to executed action—restoring accountability to social computing while preserving privacy. We argue this separation of intent (CPUX) from enforcement (infrastructure) is essential for building LLM-integrated, auditable, and socially responsible distributed systems.
CPUX, Intention Space, Design Nodes, Cognitive Computing, Data-Driven Execution, Microservices Architecture, Cloud Computing, LLM Integration, Social Computing Accountability
Oluwatobi Fadokun, Staffordshire University, United Kingdom
Cloud computing is widespread among corporations and organisations. Cloud computing offers numerous benefits, therefore many firms are switching. SME s no longer need on-site software and servers with cloud computing. IT infrastructure does not need technical experts. With the help of the cloud now businesses now rent server space anywhere in the globe to access their software applications via cloud services. With so many benefits that come with cloud computing, we have also gone into factors that can affect the adoption of cloud computing for Small and Medium Enterprises (SMEs). This article has examined a wide range of topics, from security concerns to privacy concerns to even legal considerations, that must be taken into account when a company migrates to the cloud.
Cloud, Security, SMEs, Cloud computing, Privacy, Legality.
Sairam Chennaka, Independent Researcher, India
Semantic stability—the consistency of semantic meaning across multiple generations with varying sampling parameters—is a critical but understudied property in compact language models (CLMs) with 1-4B parameters. This paper introduces the Compact Semantic Stability Framework (CSSF), a comprehensive empirical analysis framework, and proposes the Compact Semantic Stability Index (CSSI), a novel metric for measuring semantic consistency. Through extensive evaluation of five state-of-the-art models from 2025 across 9 temperature values (0.1-1.8) and 100+ languages, we analyze over 7,200 generations using more than 20 metrics. Key findings show: (1) Temperature is the dominant control factor, explaining 85.6% of semantic stability variance (ANOVA: F = 847.3, p < 0.001); (2) A fundamental stability-diversity trade-off exists (r = −0.847, p < 0.001); (3) 4B models exhibit 40% superior consistency compared to 270M baselines; (4) Semantic stability varies significantly by language resource levels (11-12% gap). Our framework provides practitioners with data-driven tools for model selection, parameter optimization, and deployment strategy. .
Semantic Stability · Compact Language Models · Generation Consistency Sampling Temperature · Semantic Embeddings
Archana Yadav and Indira Gandhi Technical University, New Delhi, India
Hinglish, a language mixture of Hindi and English is gaining popularity on Indian social media has introduced a unique challenge of sentiment analysis. Sentiment analysis of Hinglish is challenging due to inconsistent grammar, variable spelling for same word, heavy presence of slang, and emojis. This paper presents a comprehensive Hinglish sentiment classification framework using SemEval 2020 Task 9 dataset. The dataset is further expanded by adding 500 positive and 1500 negative Hinglish statements to enhance balance and diversity. Emoji-Sentiment Mapping and Slang normalization is done by designing a custom preprocessing pipeline which result in better capture of contextual meaning. Adapter-based fine-tuning of the BERT architecture has used for model training which reduce computational cost while maintaining efficient learning. The proposed model has achieved a accuracy of 64.9% and an F1-score of 0.644.To interpret model behaviour explainability modules including sentiment-wise word clouds and token-level importance visualizations, are incorporated. The study offers an interpretable and resource-efficient approach for sentiment analysis of code-mixed Hinglish text.
Hinglish Sentiment Analysis; Code-Mixed Text; Emoji-Sentiment Mapping; Slang Normalization; Adapter-Based Fine-Tuning; BERT; Explainable AI.
Parvathi Gopakumar and Vihaan Gautam, Department of Networking and Communications, SRM University, Kattankulathur, Tamil Nadu
ChoreMate is a mobile application that helps organize, assign, monitor and manage chores for all Family Members. ChoreMate allows parents to assign "Chores" to their children and/or another family member on a regular basis. As part of this process, parents are able to customize each Chore with either monetary or non-monetary rewards as an incentive for children to finish their chores and develop greater self-discipline. In addition to incorporating gamification to motivate kids to finish chores; ChoreMate also provides parents with a practical solution to help manage Family Tasks. The gamification features of ChoreMate not only help motivate kids to complete chores, they also teach kids important skills such as accountability, time management and independence. ChoreMate creates a positive and supportive home environment for parents, while giving kids the excitement and sense of accomplishment when completing their assigned tasks.
ChoreMate, gamification, chore, responsibility, track, life skills, self-discipline
Naina Jonwal¹, Atharva Joshi², Sanika Kawalkar³, Varad Khakre⁴, Nakul Sharma⁵ ¹Department of Computer Science & Engineering (Internet of Things and Cyber Security, including Blockchain Technology), Vishwakarma Institute of Technology, Pune, India
Millions of deaf and hard-of-hearing people still keep up with their visual observations and tactile senses to keep pace with what is happening around them. Even basic daily activities such as the ability to tell when one at the door is present may be challenging without effective assistive technologies. The old fashioned doorbells, which rely solely on sound, are inadequate and they are usually unsafe to persons who are not in a position of hearing this sound. This study includes a human-centered design and implementation of an IoT-based Smart Doorbell System that is designed to meet the requirements of deaf people in particular. The suggested system applies both visual alerting modes, such as bright LED signals, and vibration modules, and mobile notifications, to prevent the situation when the user misses a visitor. A microcontroller with Wi-Fi features coordinates all these components and connects with a mobile application via cloud services. The results of the testing displayed fast response rates, stable communication and high user satisfaction. This paper shows that the availability of a simple, inexpensive, and easyto-use doorbell system can benefit deaf people greatly in their independence, everyday convenience, and safety.
IoT, Smart Home System, Assistive Technology, Smart Doorbell, Deaf Assistance, Visual Alerts, Vibration Motor.
Karun Thankachan, Carnegie Mellon University, Secaucus, NJ, USA
Offline evaluation is a critical component of modern recommender system development, enabling practitioners to assess candidate models without costly and risky online experimentation. Off-policy evaluation (OPE) methods provide a principled framework for estimating the performance of new policies from historical interaction logs, and have been widely studied in the context of contextual bandits and reinforcement learning. However, real-world recommender systems typically present users with ranked lists (slates) of items, where user feedback is influenced by position bias, exposure effects, and interactions among items. Existing OPE libraries and research codebases primarily focus on single-action settings or provide limited slate support, often producing point estimates without diagnostics, uncertainty characterization, or robustness analysis. In this paper, we introduce SlateOPE, a Python library designed to address the practical challenges of offline evaluation for slate-based recommender systems. SlateOPE adopts a diagnostic-first design that treats ranked lists as first-class objects and emphasizes decision safety over point estimation alone. The library provides native abstractions for slate evaluation, plug-and-play interfaces for ranking policies, automated exposure-bias diagnostics, ranking-aware uncertainty estimation, and assumption sensitivity analysis that quantifies the robustness of evaluation results under violations of key modeling assumptions. SlateOPE further integrates simulation-based validation and explicit failure-mode detection to prevent the misuse of offline estimates in low-overlap or high-variance regimes. Unlike prior work that centers on proposing new estimators, SlateOPE focuses on unifying existing slate OPE methods within a reproducible, extensible, and practitioner-oriented framework. Through a featurelevel comparison with existing OPE libraries and research implementations, we demonstrate that SlateOPE fills critical gaps in current tooling by making uncertainty, fragility, and trustworthiness explicit. SlateOPE is intended to complement online experimentation by enabling safer and more transparent offline decisionmaking for ranking systems.
Recommender Systems, Off-Policy Evaluation, Slate-Based Recommendation, Offline Evaluation, Ranking Systems, Exposure Bias, Robustness Analysis, Sensitivity Analysis
Shudan Yue, Qingbao Li, Guimin Zhang ⋆ , and Xinyu Su, School of Cybersecurity, Information Engineering University, Zhengzhou 450001, China,
The security risk assessment of binaries has become a critical area of research in IoT firmware security analysis. Current methods for assessing binary security risks are still limited in comprehensively analyzing inter-process communication, accurately detecting vulnerabilities, and effectively evaluating risk. In this paper, we propose a security risk assessment framework for firmware binaries. Specifically, our framework first employs static analysis techniques to extract key information from the firmware, including basic firmware details, binary dependencies, component versions, and known vulnerabilities. Then, it constructs a firmware vulnerability knowledge graph containing entities, relationships, and attributes using the Cypher query language. Finally, an optimized PageRank algorithm is applied to the graph to calculate the security risk score for each binary by analyzing its dependencies within the firmware and the severity of its internal vulnerabilities. Experimental results show that the proposed framework outperforms Karonte, SaTC, and Mango in supporting inter-process communication modes and can effectively generate a list of binaries sorted by security risk, thereby assisting analysts in quickly locating high-risk binaries.
Internet of Things, Knowledge graph, Firmware binary, Risk assessment.
Ruchi Shrivastava,Research Scholar:Jiwaji University,India
Information and communication technologies (ICT) and broader information technology (IT) have increasingly become pivotal drivers of economic and social development in the 21st century. This paper explores the multifaceted role of IT in promoting productivity, innovation, governance, education, health, inclusion and empowerment, while also recognizing the associated risks, digital divides and structural challenges. Through a review of recent empirical studies and theoretical frameworks, we examine how IT contributes to economic growth and structural transformation, as well as how it fosters social development, inclusion and wellbeing. We further address the limitations and risks of technology-led development and derive policy implications, particularly for emerging and developing economies. The paper concludes that while IT offers significant opportunities for economic and social uplift, realizing those benefits depends on complementary investments in human capital, institutions, infrastructure and inclusive governance.
Information Technology, Economic Development, Social Development, ICT, Digital Inclusion, Governance, Productivity, Innovation
Ashmi S N, IIIT KOTTAYAM, India
Virtual spaces are increasingly beset with cyberbullying, which inflicts serious psychological distress. While AI moderation is available for prominent languages, low-resource languages such as Malayalam are underserved. This pa-per addresses this gap by first constructing a novel, well-balanced cyberbullying dataset of over 10,000 Malayalam comments. On this new corpus, an end-to-end detection system is developed by fine-tuning the IndicBERT model for multi-class classification. Experimental results establish a strong performance benchmark, with the model achieving 93% accuracy. Furthermore, the application of LIME and SHAP provides crucial interpretability, pinpointing the specific token-level drivers behind the model’s predictions and uncovering its underlying biases.
Cyberbullying Detection ∙Malayalam NLP ∙Explainable AI ∙LIME ∙SHAP∙Content Moderation
DAVID, PATRICK ONUWABUCHI, Alex Ekwueme Federal University Ndufu-Alike Ikwo.
The increasing digitization of data across healthcare, financial services, and other critical sectors has amplified the importance of safeguarding information privacy and ensuring robust digital security. This paper provides a comprehensive examination of the conceptual, technical, and regulatory aspects of information privacy and digital security, highlighting the interplay between technological solutions, human factors, and ethical considerations. Key areas explored include cryptography, access control, privacy-enhancing technologies, and artificial intelligence-driven security mechanisms, alongside sector-specific challenges illustrated through case studies in healthcare and financial services. The study also investigates regulatory frameworks, including GDPR and CCPA, and addresses compliance and ethical challenges in data collection and processing. Finally, future trends, emerging technologies, and recommendations for research are discussed to guide the development of resilient, privacy-aware, and secure digital ecosystems. The findings underscore that a multi-dimensional approach integrating technical innovation, regulatory compliance, ethical governance, and user awareness is essential for effective protection of sensitive information.
Information Privacy, Digital Security, Privacy-Enhancing Technologies (PETs), Cybersecurity, Regulatory Compliance, Privacy-by-Design, Security-by-Design, Artificial Intelligence.
Jin Jin *, Xiaoyong Zhang, Jun Li, and Mengce Zheng, School of Information and Intelligent Engineering, Zhejiang Wanli University, Ningbo, Zhejiang, China
With the surge in the number of college graduates, the assessment of the employment status of college graduates has become particularly important. By deeply mining student grades and employment data, colleges and universities can improve talent training programs, innovate curriculum settings, and enhance student employability. This study is dedicated to using machine learning methods to predict the employment status of college graduates through student grades. Specifically, we collected and sorted out the academic performance and employment data of information majors in a certain university in the past few years, and then divided the courses into six categories according to the talent training program and course attributes to make a standard data set for public use. Then, we used database technology to calculate the average grade point of each course, and then used six machine learning methods to predict and analyze the employment status of graduates. We divided these data into training sets and test sets and used five-fold cross-training. The average accuracy of five-fold cross-validation on some models can reach more than 90%. By comparing and analyzing these methods, we also found some key indicators that have a significant impact on the employment prospects of college students, namely quality development courses and total grade points.
College student employment; Data mining; Artificial intelligence; Deep learning