Autism spectrum disorders (ASD) encompass a range of complex neurodevelopmental conditions characterized by challenges in communication, social interaction, and the presence of repetitive behaviors. The early and accurate diagnosis of ASD is crucial for optimizing outcomes and enhancing the quality of life for individuals and their families. The World health organization (WHO) reports that approximately 1 in 100 children are affected by ASD. Traditional diagnostic methods, such as the Autism diagnostic interview-revised (ADI-R) and the Autism diagnostic observation schedule (ADOS-2), provide structured and reliable evaluations. However, integrating advanced techniques such as Opinion mining (OM) from Machine learning (ML) framework offers promising enhancements to current diagnostic practices. OM leverages Natural language processing (NLP) to analyze textual data and discern human perceptions and emotions, potentially uncovering subtle indicators of ASD that standard methods may overlook. In this research, we propose an innovative diagnostic model combining OM with ML classifiers, including Support vector machines (SVM), K-nearest neighbors (KNN), and Naïve Bayes (NB). The results indicate that the KNN classifier exhibits superior accuracy with 88.78% in category 2, SVM showed the second-best performance with an accuracy of 86.75% in category 1, and NB had its best result in category 3 with an accuracy of 83.44%. This interdisciplinary approach underscores the value of integrating scientific advances with clinical experience, paving the way for more accurate and earlier detection of ASD.