Cube Index for Text and Image Information Retrieval and Mining

This paper proposes a novel cube index for text and image information mining. Based on the study of text and image index used for information retrieval, this paper proposes that information mining on unstructured text and image data is possible from the index structures that exist already. The cube index merges the text and image index proposed earlier, into a single structure. This mimics the brain in its capacity to view the text and image data simultaneously and retrieve them together, according to the query need of the user. It will help in the indexing, retrieval and mining of both text and image data present in the document database. The online analytical processing (OLAP) over the cube index and concept hierarchy for the document and image are proposed. The operations that can be performed on the data cube are applied to the cube index to help with the retrieval and mining. The study gives a good guarantee that the model can be used for database analysis of text and image, based on the results from the retrieval experiment.