Paper Title
Reduction of The Semantic Gap Using Pseudo Relevance Feedback Algorithm

With the exponential growth in the Internet Technology, Image analysis and retrieval is a typical domain and high degree of abstraction is required for which semantic gap is major key problem. To overcome these issues, people need to organize and access the increasing amount of user-tagged multimedia. A tag gives the descriptive information of an image so it becomes necessary to incorporate both image content and tag information for image retrieval and image annotation task. The proposed technique utilizes global low level features of an image such as color, texture, pattern and edge and build similarity graph to find nearest neighbors of an image. Finally, A scalable and novel random walk for fusing the parameters to balance the visual content and tag. This Proposed System provides the improvement in the retrieval performance of Content Based Image retrieval, Text based image retrieval and image annotation.