aspg-21-3114-1791475512 20261008160512 American Scientific Publishing Group admin@americaspg.com American Scientific Publishing Group International Journal of Neutrosophic Science IJNS 2692-6148 2690-6805 2025 25 2 Abdalla Ibrahim Abdalla Musa Department of Computer Science, College of Computer, Qassim University, Buraydah, Saudi Arabia Mohammed Abdullah Al-Hagery Department of Computer Science, College of Computer, Qassim University, Buraydah, Saudi Arabia Sentiment Analysis (SA) is a crucial task for analyzing online content over languages for processes such as content moderation and opinion mining. However advanced NLP modeling approaches frequently need an abundance of training datasets to accomplish their outcomes. SA is a classification task where the polarity of text dataset is detected, viz., to analyze a document or sentence expressing a positive, negative, or neutral sentiment. Deep learning (DL) becomes predominant in resolving Natural Language Processing (NLP) tasks. On the other hand, this technique requires a significantly enormous quantity of annotated corpus, which is not easier to attain, particularly under these lower resource settings. Neutrosophic Net-RBF Neural Network (NNRBFNN) combines the principle of neutrosophic logic (NL) with RBF-NNs for handling data indeterminacy and uncertainty. This combined strategy optimizes conventional NNs by incorporating the possibility of addressing incomplete and imprecise data, augmenting decision-making in challenging circumstances. This paper introduces a Neutrosophic Net-RBF Neural Network with Sentiment Analysis on a Low Resource Language (NNRBFNN-SALRL) model. To accomplish this, the NNRBFNN-SALRL method undertakes data pre-processing to transform the input dataset into a helpful format, and Term Frequency Inverse Document Frequency (TF-IDF) technique is utilized for the process of word embedding. For the classification method, the NNRBFNN model is used. To optimize the recognition outcomes of the NNRBFNN method, the hyperparameter tuning technique can be done using the Bayesian Optimization Algorithm (BOA). Wide-ranging experiments were conducted to validate the superior outcomes of the NNRBFNN-SALRL method. The empirical findings indicated that the NNRBFNN-SALRL method emphasized betterment over other approaches. 2025 44 56 3114 https://creativecommons.org/licenses/by/4.0/ 10.54216/IJNS.250205 https://www.americaspg.com/journal/21/article/3114 https://www.americaspg.com/storage/51737211005.pdf