Cross-lingual Related Events Recognition Methods Based on The Event Central News Sets

To help the public fully understand the ins and outs of an international or regional major event, we propose two cross-lingual related events recognition methods to acquire more related events in the news on different countries’ internet. Firstly news in different languages that focus on a same event are clustered into a news set. Then two method will be used together to determine whether two events are related. One is explicit related event recognition method which determines that by checking whether there is a description of one event can be found from another event central news set. The other one is implicit related event recognition method which determines that by a SVM classifier with several correlation and similarity features of two events. The experimental results show that the two methods are complementary to each other, and it is practical to use them together.