Online social networks have introduced diverse discussions regarding different topics. At the same time, these platforms have also garnered the idea of shaping opinions. The increase in user engagement in case of arguments has led to the issue of polarization in online social media platforms. In this paper, we have studied the influence of opinion change on homophily in social networks and the variation in polarization. Our own new method has been proposed for analysing the modification of user opinions dynamically in different timestamps. The probability of information sharing varies depending on the opinion similarity and homophily in social networks alters accordingly. Our approach portrays the opinion modification dynamics for different network features using complex networks concepts. We have observed that on social interaction networks like Twitter, opinion, information sharing and homophily perform a significant role in the determination of online polarization. Our approach has been compared with an established approach over external datasets and the results obtained are promising.