A sentiment analysis hybrid approach for microblogging and E-commerce corpus

摘要: Exploiting linguistic features is necessary on sentiment analysis in natural language processing. This paper proposes a novel approach exploiting and SVMperf based semantic classification. The innovation that it uses the dependency relationship to do feature extraction. In order reduce computational complexity, this X2 (chi-square) Pointwise Mutual Information (PMI) metrics for selection. Furthermore, as analysis, algorithm alternative structural formulation of SVM optimization problem two different corpuses (i.e., microblogging e-commerce data set) evaluate performance. Experiment results show feasible approach. Existing problems further works are also present end.