Many real-life systems are framed as networks which are used to model interactions between complex entities. However, in many scenarios these interactions may not be pairwise, rather should described as higher-order interactions. For such a scenario Hypergraphs are preferred rather than simple Laplacian. Therefore, to decide whether Laplacian, Hypergraph or both could be used for the given system could lead to a problem which needs to be addressed properly. In this paper, we present a semi-supervised framework which considers weighted combination of Hypergraph and Laplacian information for pattern classification. Numerical experiments on seven binary, five multi-class and four multilabel datasets along with MNIST-fashion dataset validate the efficacy of proposed algorithm.