Cross-media topic detection method and device based on multimodal information fusion and graph clustering

Abstract

The invention discloses a cross-media topic detection method and device based on multimodal information fusion and graph clustering. The method includes the steps that first, text information and video content are preprocessed; second, text fusion and high-layer characteristic extraction are performed to obtain text information similarity; third, similar clips of video are detected, and vision information similarity is obtained; fourth, according to the text information similarity, the vision information similarity and time information of fused data, the final data similarity is calculated; fifth, according to the final data similarity, graph fusion and graph clustering are performed, and topic detection is completed. The method effectively avoids the over-segmentation and the over-generation problem caused by hard quantification on a time axis and the problem that an existing topic detection method can not be transplanted to topic detection on multimodal data from different multimedia sources.

Description

technical field [0001] The invention relates to a topic detection technology in multimedia data, in particular to a cross-media topic detection method based on multimodal information fusion and graph clustering. Background technique [0002] The rapid development of social networks and social media, as well as the high-speed dissemination of data on social networks and social media, make topic detection from network data one of the research hotspots in the field of information processing. Reasonable and effective organization of network data according to topics, on the one hand, it can facilitate users to browse network data, improve user experience, so that they can find the social reality topics they want to know more quickly and accurately, on the other hand, it is beneficial for network administrators and government-related department for information management. [0003] At present, most methods for topic detection can be roughly divided into three categories according ...

AI Technical Summary

Problems solved by technology

[0010] The technical problem to be solved by the present invention is to provide a cross-media topic detection method and its device based on multi-modal information fusion and graph clustering, which are used to effectively avoid over-segmentation and over-generation caused by hard quantization of the time axis. problem and the problem of topic detection where existing topic detection methods cannot be ported to multimodal data from different media sources

Images