@Article{iasc.2023.033915, AUTHOR = {Najm Alotaibi, Badriyya B. Al-onazi, Mohamed K. Nour, Abdullah Mohamed, Abdelwahed Motwakel, Gouse Pasha Mohammed, Ishfaq Yaseen, Mohammed Rizwanullah}, TITLE = {Political Optimizer with Probabilistic Neural Network-Based Arabic Comparative Opinion Mining}, JOURNAL = {Intelligent Automation \& Soft Computing}, VOLUME = {36}, YEAR = {2023}, NUMBER = {3}, PAGES = {3121--3137}, URL = {http://www.techscience.com/iasc/v36n3/51899}, ISSN = {2326-005X}, ABSTRACT = {Opinion Mining (OM) studies in Arabic are limited though it is one of the most extensively-spoken languages worldwide. Though the interest in OM studies in the Arabic language is growing among researchers, it needs a vast number of investigations due to the unique morphological principles of the language. Arabic OM studies experience multiple challenges owing to the poor existence of language sources and Arabic-specific linguistic features. The comparative OM studies in the English language are wide and novel. But, comparative OM studies in the Arabic language are yet to be established and are still in a nascent stage. The unique features of the Arabic language make it essential to expand the studies regarding the Arabic text. It contains unique features such as diacritics, elongation, inflection and word length. The current study proposes a Political Optimizer with Probabilistic Neural Network-based Comparative Opinion Mining (POPNN-COM) model for the Arabic text. The proposed POPNN-COM model aims to recognize comparative and non-comparative texts in Arabic in the context of social media. Initially, the POPNN-COM model involves different levels of data pre-processing to transform the input data into a useful format. Then, the pre-processed data is fed into the PNN model for classification and recognition of the data under different class labels. At last, the PO algorithm is employed for fine-tuning the parameters involved in this model to achieve enhanced results. The proposed POPNN-COM model was experimentally validated using two standard datasets, and the outcomes established the promising performance of the proposed POPNN-COM method over other recent approaches.}, DOI = {10.32604/iasc.2023.033915} }