Heterogeneous tag fusing system based on semantic structure feature analysis

Abstract

The invention discloses a heterogeneous tag fusing system based on semantic structure feature analysis. Sorting and fusing of different tag systems of various industry websites are concentrated on, and the heterogeneous tag fusing system based on the semantic structure feature analysis is put forward, and is used for constructing fine-grit and structuralized tag systems. Since tag systems coming from different sources have different structures, tag webpage information is utilized to fuse the tag systems, and by identifying the mapping relation, namely the equivalence relationship and the hyponymy relationship, of two tag systems, heterogeneous tag systems are merged. Since the tag systems from a mainstream website are selected, it is guaranteed that the tag systems are more authentic and more persuasive in source, so that the obtained tag systems are more refined and accurate, and a basis is laid for user behavior portrayals.

Description

technical field [0001] The invention belongs to the technical field of big data applications, and in particular relates to a heterogeneous tag fusion system based on semantic structure feature analysis, which lays a foundation for user behavior portraits by constructing a tag system. Background technique [0002] With the in-depth research and application of big data technology, enterprises are increasingly focusing on how to use big data to serve precision marketing, and then deeply tap potential business value. As a result, the concept of "user portrait" came into being, laying the cornerstone of the big data era and providing sufficient information foundation for enterprises. At present, user portrait technology has been widely used in advertising marketing and customer loss analysis and other fields, such as recommendation systems and purchase predictions. User portraits, that is, user information labeling, means that companies can perfectly abstract a user's business p...

AI Technical Summary

Problems solved by technology

For the labeling system built by the collaborative labeling system, there is no obvious hierarchical structure, and there are repeated labels; for the labeling system built by the ontology method, the main reason is that the ontology construction method itself is not mature and requires a lot of manual participation; for the topic model extraction The label system constructed by this method has relatively inconspicuous hierarchical structure, and there is a problem of coarse-grainedness, that is, the label is too broad to accurately describe user behavior characteristics

Images