Time line inference from story sentence with time series

Recently, we can find many literary works on the Web. Automatic summarization is effective method to find favorite work through various works. But, stories have no logical structure like minutes or news papers. So, many exiting researches of automatic summarization are not effective for stories because these target minutes or news papers only. System optimized for stories is necessary. We focused on time series at story, a type of structure of story. We organized the system to infer it. System infers four types of time series by using timeword database. We tested features of four time series. System was checked it worked at story sentences.