Towards a distributed federated architecture for digital documents

Creating the federated architecture is the most significant issues in the field of digital library. Human perception is not uniform while measuring the relevance to automate the retrieval process. In this work we have designed a system for integrating the existing architectures for digital library. This architecture uses integrated systems such as metadata, standard descriptors, feature extraction etc for text searching and retrieval. Databases of different size were used to estimate the accuracy of the system. The proposed algorithm works on the concept of minimum weight tree that removes the irrelevant texts from the retrieved hits, based on the dynamic threshold provided to the algorithm. We found out that careful combination of the different features based on our proposed heuristic, can increase the creation of a unified architecture for digital libraries.