Summarization based on multiple feature combination

This paper presents the state of art research progress on Chinese single document summarization. Our method utilizes hLDA(hierarchical Latent Dirichlet Allocation) algorithm to model the documents first. A new feature is proposed from the hLDA modeling results, which can reflect some semantic information. Then it combines different traditional features to extract summary sentences from the documents. Four kinds of sentence extraction algorithms were designed to explore the effect of hLDA clusters. Our method attempts to find an optimal linear feature combination for sentence scoring through large scale experiments. After the comparison with other summarization methods, our method reveals better performance.