UCLA

There is a ton of news generated every day. However, it is more than anyone can analyze. Since reading every single news is impossible, presenting key information extracted from news can highly improve the efficiency for accessing massive knowledge pieces. And people nowadays analyze news by applying various kind of natural language processing technique both of linguistic approach and machine learning approach. Different methods can interpret news in different ways. Various natural language processing approaches, such as semantic role labeling and coreference resolution, have been applied in information extraction and we applied the technique for analyzing news. We extracted global entities information from news and presented a demo to visualize different aspects of news data. To better extract key information from massive news, we focused on semantic roles in the news data, and the relationship between them in the news flow.