In machine translation (MT), one of the challenging tasks is to translate the proper nouns and technical terms from the source language to the target language while preserving the phonetic equivalent of original term. Machine transliteration, an essential part of MT systems, plays a vital role in handling proper nouns and technical terms. In this paper, a hybrid attention-based encoder–decoder machine transliteration system is proposed for the low-resource English to the Assamese language. In this work, the proposed machine transliteration system is integrated with the previously published hybrid attention-based encoder–decoder neural MT model to improve the translation quality of English to the Assamese language. The proposed integrated MT system demonstrated good results across various performance metrics such as BLEU, sacreBLEU, METEOR, chrF, RIBES, and TER for English to Assamese translation. Additionally, human evaluation was also conducted to assess translation quality. The proposed integrated MT system was compared with two existing systems: the Bing translation service model and the Samanantar Indic translation model.

Tokenization is a fundamental step in NLP, influencing both computational efficiency and downstream task performance. While subword methods such as Byte-Pair Encoding (BPE), WordPiece, and Unigram have shown strong results for high-resource languages, their suitability for low-resource and morphologically rich languages like Assamese remains insufficiently understood. This study presents a systematic evaluation of these tokenizers on a curated Assamese Wikipedia corpus, examining intrinsic efficiency metrics—including subword fertility, compression ratio, tokenization speed, and token diversity—alongside statistical validation and energy trade-offs. We further connect intrinsic behaviour to practical outcomes by fine-tuning IndicBERT and mBERT on sentiment and hate-speech tasks, and by assessing morphological boundary preservation. Results show that BPE-32K provides the most compact and semantically coherent segmentation, improves downstream F1 scores by 3–4 points, and preserves morpheme boundaries in 82% of cases, while WordPiece-16K offers the fastest tokenization speed. Overall, the findings demonstrate that vocabulary scaling reduces over-segmentation and that language-specific tokenizers substantially outperform multilingual defaults for Assamese. This work provides empirical guidelines for selecting tokenizers tailored to low-resource Indic languages and for integrating them effectively into LLM pipelines.

NMT for Assamese language remains challenging because of its morphological rich language and lack of parallel resources. The paper aims to study the effect of word embeddings, different neural architectures and bidirectional pre-training for enhancing bidirectional English–Assamese translation. Three embedding methdos (Word2Vec, GloVe and FastText) were evaluated within RNNs using GRU and LSTM architecture. FastText consistently yielded the highest BLEU scores among recurrent models, achieving 24.99 for Assamese–English and 24.65 for English–Assamese, owing to its ability to capture subword level information and handle rare word forms. To establish stronger baselines, we fine tuned Transformer-based models (MarianMT and mBART) on a 175,000-sentence parallel corpus and further enhanced mBART through a Bidirectional Training (BiT) phase to promote cross-lingual parameter sharing. The BiT-pretrained variant achieved the best performance, with BLEU scores of 28.93 (As–En) and 28.57 (En–As). Finally, we compared these systems against a zero-shot GPT baseline, which underperformed relative to the fine-tuned models, reaffirming the effectiveness of specialized training for low-resource translation. Human evaluation of fluency and adequacy supported these results, confirming that Transformer-based architectures with bidirectional pre-training remain the most effective solution for English–Assamese NMT.

Industry 5.0 places emphasis on human-centered innovation, aligning digital transformation with sustainability, resilience, and societal well-being. Smart cities serve as practical environments where advanced technologies must balance operational efficiency with environmental responsibility. Since October 2023, deep neural networks have been widely used in urban decision-making, including traffic optimization and environmental monitoring. However, the high energy demand required to train large-scale models has raised concerns regarding both environmental impact and economic cost, leading to growing interest in sustainable AI practices. This study explores Carbon-Aware Neural Network Optimization within the Green AI framework as a practical approach to reduce computational carbon emissions while maintaining predictive performance. Three complementary strategies are integrated into a unified training pipeline and monitored using the CodeCarbon framework: partial layer freezing, quantization-aware training, and adaptive early stopping. Experiments are conducted on CIFAR-10 using MobileNetV2 and on a subset of ImageNet with ResNet-50. Results show a reduction in CO₂ emissions of 52% and 49%, respectively, with only minor changes in model accuracy. The findings indicate that combining these techniques provides an effective balance between efficiency and predictive reliability, supporting the development of sustainable AI solutions for Industry 5.0 and smart city applications.

People across the globe are actively participating in online groups due to excessive usage of social media as channel of communication and data exchange. Also, due to social media platforms such as Instagram, Facebook social communication has seen a considerable increase. Social media's emergence has revolutionized how people talk to one another and engage with the world. However, there is mounting apprehension regarding the potential repercussions of social networking sites on mental welfare. This paper examines the effect of social networking sites on mental health from a social informatics perspective. Drawing on theories and concepts from social informatics, this paper explores the social, cultural, and technological factors that influence the connection between social networks consumption and psychological health. The paper discusses the possible risks and benefits of social media usage and offers commendations for designing social media platforms that promote psychological health and wellbeing.