Semantic annotation for hybrid contents

As computer technology becomes more advanced and information exchange grows into faster and easier, huge amount of digital data has emerged as a leading role of the information era. Through the social networking service, Internet users widely post or discuss the interests via blogs about music, journey, experience of products and so on. An article could be treated as an Interest resource containing hybrid contents, such as texts, images, video, etc. However, the article is just like an isolated island. No relations are created between any two articles. To build the connections, mining the useful data or images in the boundless articles one by one is a time-consuming process. To resolve the problem, the paper provided a semantic annotation mechanism for the hybrid contents, which are benefit that the articles usually contain abundant images to enrich the contents and the texts are helpful to annotate the images.