Nowadays, social networking sites such as Facebook, Twitter, LinkedIn, YouTube, and other e-commerce websites produce a large number of text reviews. These text reviews mostly describe the product features and their opinions, which are the most important to the product developers, launchers, or buyers for business development and decisionmaking processes. Therefore, we present an opinion-based co-occurrence network for product reviews. The main aim of this research is to identify the popularity of product features or popular terms, the number of connections of a term, the strong relationship between terms, grouping the product terms, and the sentiment polarity links between terms in both positive sentiment and negative sentiment. Also, we employed the Harel-Koren fast multiscale layout algorithm and CNM (Clauset-Newman-Moore) algorithm for visualizing and grouping the network. We then measured the overall graph metrics and vertex metrics to characterize the network. Additionally, the experimental result shows the ranked product features and their social strength between product features and sentiments.