Public opinion event recommendation method, system and terminal combining content and atlas

Abstract

The invention belongs to the technical field of network space cognition, and discloses a public opinion event recommendation method, system and terminal combining contents and maps. The method comprises the following steps: introducing a domain knowledge graph into related public opinion event recommendation, integrating public opinion events into existing domain graph data by adopting an entity extraction and relationship extraction knowledge graph construction method, and calculating knowledge representation of the events by adopting a TransE knowledge embedding method; and for the obtained final event representation, taking the influence index of the public opinion event as additional supplementary information, and recommending k-n events with the maximum influence value for the events of which the recommendation number does not meet k as supplement. According to the method, the domain knowledge graph is innovatively introduced into related public opinion event recommendation, public opinion events are integrated into existing domain graph data by adopting knowledge graph construction methods such as entity extraction and relation extraction, and knowledge representation of the events is calculated by adopting a TransE knowledge embedding method.

Description

technical field [0001] The invention belongs to the technical field of cyberspace cognition, and in particular relates to a public opinion event recommendation method, system and terminal combining text content and knowledge graph. Background technique [0002] At present, with the development of technology, the Internet has become the main carrier of public opinion fermentation and dissemination, and public opinion events mainly refer to events that arouse netizens' discussion, attention and dissemination on the Internet. [0003] In the absence of any user behavior data, the available recommendation methods are still mainly content-based related item (event) recommendation models. Existing technical solutions use language models to characterize the text semantics of event content, and use different similarity calculation methods to calculate the similarity between current event representations and historical event representations, and finally input the k most similar relat...

AI Technical Summary

Problems solved by technology

[0011] Based on the problems existing in the existing public opinion event recommendation methods, this scheme proposes an event recommendation method that combines knowledge representation and text content representation, introduces external information from domain knowledge graphs, and takes into account the complex relationships behind different events. Solved the problem of being unable to describe the deep semantics and domain characteristics of event text

Images