Polarity classification of opinionated sentences with both positive and negative senti-ments1 is a key challenge in sentiment analysis. This paper presents a novel unsuper-vised method for discovering intra-sentence level discourse relations for eliminating polarity ambiguities. Firstly, a discourse scheme with discourse constraints on polarity was defined empirically based on Rhetorical Structure Theory (RST). Then, a small set of cue-phrase-based patterns were utilized to collect a large number of discourse instances which were later converted to semantic sequential representations (SSRs). Finally, an unsuper-vised method was adopted to generate, weigh and filter new SSRs without cue phrases for recognizing discourse relations. Experimental results showed that the proposed methods not only effectively recognized the defined discourse relations but also achieved significant improvement by integrating discourse information in sentence-level polarity classification.