Abstract: In the era of social media, the innovative communities(IC) have become important habitats of lead users. Therefore, how to identify leading users in the big data environment is a key issue for companies to capture value from innovation communities. In this paper, we propose a novel approach that relies on extracting linguistic-style cues from community posts to identify lead users. Text mining were used to extract positive emotional, collectivism and self-interest-oriented writing style cues from 47310 posts written by 355 members of 10 ICs. The results show that positive emotional and self-interest-oriented linguistic-style are significantly positive related to lead-userness, they can play signaling roles for identifying lead users. While, the relationship between collectivism linguistic-style and lead-userness are not significant. The theoretical and managerial implications of this novel approach of identifying lead users were concluded.