Chinese Text-Oriented Interpersonal Emotional Relationship Mining Based on Event Knowledge Model

The extraction of emotional relationships between different persons in text is important for the study of social networks. Most current interpersonal relationship extraction methods focus on the extraction of explicit and direct interpersonal relationships text, ignoring indirect or even uncertain emotional relationships hidden in the text. In this paper, we propose an interpersonal emotional relationship mining algorithm based on event knowledge model. The event knowledge model describes the semantic relationship between different event classes, which are related different types of participants. Therefore, the event knowledge model contains the relationship between different roles in a specific event scenario. Based on the semantic relationship between events and event roles in event knowledge base, the relationship between persons involved in events could be inferred, and the emotional attitude of persons could be obtained by classification learning. Experiments show that our method has a greater effect than traditional relationship mining methods.