Knowledge-fused social network streaming event detection system

The invention discloses a knowledge-fused social network streaming event detection system, and the method comprises the steps: extracting a plurality of types of keywords from event data, wherein theplurality of types of keywords and an event form nodes in a heterogeneous network; selecting a plurality of meta-paths from the heterogeneous network, and performing randomly walking in the heterogeneous network according to the meta-paths to obtain paths as corpora; inputting the corpus into a Streaming light RNN for training, and obtaining embedded representations corresponding to all the nodes;and calculating a similarity between events based on the embedding representation. 本申请公开了一种融合知识的社交网络流式事件检测系统，所述方法包括：从事件数据中提取多种类别的关键词，其中，所述多种类别的关键词以及所述事件本身构成异构网络中的节点；从所述异构网络中选择多种元路径，将在异构网络中按照各种元路径进行随机游走得到路径作为语料；将所述语料输入到Streaming lightRNN中进行训练，得到各个节点对应的嵌入表示；基于所述嵌入表示计算事件之间的相似度。