Research based on concept maps and Hidden Markov Model for multi-document summary

A method of generating multi-document summary was proposed based on concept maps and Hidden Markov Model. Handling the original text file, extract concept based on improved Hidden Markov Model and generate concept map, finally calculate the importance and related degree of sentences according to concept map, and then generate semantic summary. The experiment result shows that the summary generated by this method is better than method based on statistics in readability.