Summary: | 碩士 === 國立雲林科技大學 === 資訊管理研究所 === 87 === In this information exploding age, we have to spend more and more effort and time to find out useful information from huge amount of text documents using traditional methods such as full-text retrieval or keyword retrieval. Espe-cially, extracting important information from free form documents and finding out the relationship among extracted information are even harder. Once the information is extracted, it may help in critical decision-makings. The pur-pose of this study is focused on extracting information from electrical social news on internet. The information extracted in this study includes person names, event, time, places and things in news articles. The extracted informati-on can be further used in news data mining or tracking related news reports.
|