Exploiting sentiment relations to improve the accuracy of sentiment analysis has caught the interest of recent research. When expressing their opinions, users apply different sentence syntactic constructions styles. This analysis leverages on a sentiment lexicon that includes general sentiment words that characterize the overall sentiment towards the targeted named-entity. However, in most cases, target entities are themselves part of the sentiment lexicon, creating a loop from which it is difficult to infer the overall sentiment to the target entities. We propose the application of conditional random fields (CRF) to predict opinion target labels. More specifically, we exploit a set of opinion patterns to extend an opinion word lexicon and then propose to apply a CRF algorithm to detect the interactions between opinion expressions and opinion targets.