Abstract
Automatic text summarization is to summarize documents automatically via a computer system. Automatic text summarization can help readers to save time in information collection. Conventional text summarization methodologies do not take reader’s individual requirements into consideration. In order to generate document summaries that can meet readers’ individual requirements, this research develops an automatic document summarization model that generate summaries based on readers’ individual requirements. To establish this summarization model, the document summarization problem is transformed into a mathematical optimization problem by analysis of the quality factors for summary and calculation of summary quality indices and constraints of quality factors. The genetic algorithm can be applied to solve the mathematical optimization problem for text summarization, and then the text summary can be generated according to the optimal solution acquired by the genetic algorithm. On the basis of the proposed methodology, this research develops an automatic text summarization system that generates summaries according to user’s individual requirements. To evaluate the performance of this system, this research collects journal papers as testing data to test system performance. In order to evaluate system performance, this research compared the summaries written by authors with thoes generated by the proposed system. The experiment results show that this text summarization system not only well summaries documents, but also appropriately take reader’s individual requirements into consideration.