With the Web 2.0 paradigm, users play the active roles in producing Web contents at online forums, wiki, blogs, social networks, etc. Among these users contributed contents, many of them are opinions about products, services, or political issues. Accordingly, extracting the comparative relations about products or services by means of opinion mining techniques could generate significant business values. From the producers’ perspective, they could better understand the relative strength or weakness of their products, and hence developing better products to meet the consumers’ requirements. From the consumers’ perspective, they could exercise more informed purchasing decisions by comparing the various features of certain kind of products. The main contribution of this paper is the development of a novel Support Vector Machine (SVM) based comparative relation map generation method for automatic product features analysis based on the sheer volume of consumer opinions posted on the Web. The proposed method has been empirically evaluated based on the consumer opinions crawled from the Web recently. Our initial experimental results show that the performance of the proposed method is promising, and the precision can achieve 73.15%.