Multi-document Summarization Based on Word Feature Mining

This paper discusses an approach to multi-document summarization that builds on understanding word as feature deeply. We created 7 basic word features using the frequency, position information, event information and topic information. Then choose logistic regression model to compute words value. The summarizer gives a score of sentence by words value, and combines score and redundancy of sentence to produce summarization. The evaluation of summaries uses three parameters which are N-gram co-occurrence statistics, term word coverage and high frequency word coverage. The experiment results show the systempsilas has more effectiveness and feasibility.