Multi-dimensionality content labeling method based on semanteme label database

Abstract

The invention discloses a multi-dimensionality content labeling method based on a semaneteme label database. The multi-dimensionality content labeling method based on the semaneteme label database comprises the steps of establishing the semaneteme label database, configuring extensible resource types, establishing a multi-level configurable content labeling dimensionality, dividing dimensionalities of resources according to content features to establish a multi-level content dimensionality, establishing corresponding relationship between the configurable and revisable resource types and the content labeling dimensionality, conducting resource content labeling based on the semaneteme label database, processing temporary labels, conducting resource search based on the semaneteme label database, inputting index words by a user, and conducting automatic match in the extensible database by the system. If the match is successful, the system searches corresponding images according to corresponding label labeling codes. If the match is unsuccessful, the system can match the index words with resource description information, and at the same time the system saves the index words into a temporary label database. Precision and efficiency of resource labeling are effectively improved, and good base is founded for resource search and data analysis.

Description

technical field [0001] The invention relates to the fields of data mining, data analysis and knowledge reasoning, and designs and implements a multi-dimensional, semantic and structured labeling method for resource content. Background technique [0002] In recent years, with the rapid development of economy and society, the number of resources has increased rapidly, but the development of resource labeling has been relatively slow, and the problem of resource retrieval has become increasingly prominent. Research on resource labeling methods can effectively solve the problem of resource management and retrieval, improve the utilization rate of resources, and meet the requirements of resource efficiency, use and management. Reasonable and efficient utilization plays a positive role in promoting. [0003] At present, there are many resource labeling methods, which can be mainly divided into resource attribute labeling methods, resource content-based feature labeling methods, r...

AI Technical Summary

Problems solved by technology

This method connects the original isolated resources and increases the coupling between different resources. The resource ontology provides a formal basis for standardized labeling, and the marked resources correspond to the domain ontology, which can realize the intelligent retrieval of resources. ; However, the construction of domain ontology cannot be completed overnight, and resources involve a wide range of categories, and it is currently not practical to completely rely on ontology for general and complete labeling of resources

Images