Social media platforms have become an easy method of communication for many users. Content posted on social media can influence those who are exposed to it, and users who posted that content are referred to as influencers. Identifying influencers has many applications in marketing, politics, and even health awareness. While research identifying influential users across multiple fields has been studied extensively, users’ influence varies in different topics. Recent studies in topic-specific influence have shown that identifying influencers on the topic-level is more effective. However, most of the existing influencer detection approaches focus only on influential user identification and do not consider that some content can be influential regardless of who published it. This paper investigates the problem of detecting topic-specific influential users and tweets in Twitter datasets. We introduce HyperTwitter, a framework that uses a Twitter sub-graph consisting of users, tweets, and interactions as input. HyperTwitter generates a hypergraph with hyperedges of two types: networks and topic edges, then measures the topic distribution for both users and tweets. With this distribution and the constructed hypergraph, we create a local, topic-based influence ranking for each user and tweet. We conduct extensive experiments with two Twitter datasets and show that the proposed framework outperforms existing baselines significantly.