The Cross-domain Model Of Learning Based On Co-algebras And Application

Across Domains Learning is change from the traditional migration on learning,Is designed to construct a model state structures with mathematical Co-algebras between multimedia data, To finding some models and features more in line with the data structure of the relevance of a more stable and establish a more stable structure and associated diagram for multimedia data, but due to the direct data demonstrated between similarity and relevance vary widely, it is possible to find some common features represent different media represent the most important part of multimedia search, if we can solve this problem, the amount of data types and the data utilization so wide of the case will have a tremendous effect of promoting in the search for a variety of multimedia representations are made some results, but still not up in the structure of data has proved between similarity and relevance, in view of this situation, we research a learning-based and cross-co-algebra homomorphism,sift feature extraction and multimedia:First prove Co-algebras meet over induction and as the biggest advantage over the inductive method then an open structure, the nature of the interaction can maximize search results leading to more diverse enough search model, the model based on the demand predetermined threshold value All search data thresholds are met in the multimedia retrieval is very common and to find the common features between the multimedia data representation in cross-border areas between the multimedia learning, especially in areas of high similarity, cross-domain learning to get the best effect of the transition effect.Secondly, using a high degree of similarity between the multimedia data to building a structural connection between FIG multimedia,and established between text, images, audio, video search Fig association in order to attempt to search map to FIG text search, image, video and other multimedia files related to higher and better in one to one data migration between many transfer to meet the diversity of current multimedia search.