An enabled pile clustering and collaborative community sifting for information and image retrieval system

Abstract

The main objective of this research work is to develop the system for retrieving information and images from web pages. It provides challenging solutions to the industry experts and researchers due to overcrowded information. Millions of people use web search engines every day to carry out huge volume of tasks, for instance, from shopping to research. In web searching, the information retrieved by the user is not always appropriate. It provides ambiguous information for the raised query which leads to the situation where the user cannot get relevant information within the stipulated time. This work is developed to overcome these issues for retrieving the information and images according to the user queries using Enabled Pile Clustered exact content Retrieval and Repository (EPCRR) and the Collaborative Community Oriented Sifting (CCOS) procedures. In the first approach, pile clustering and automated collaborative filter are used where it provides the exact information to the user. The second approach provides the relevant images using both visualization and documented information. In EPCRR, the data set is formed and the functional and description similarities are calculated using Jaccard Similarity Coefficients (JSC). The characteristics similarity is calculated using functional and description similarities. The user and item profiles are collected and clustered using the hierarchical agglomerative clustering algorithm. The rating similarity of each cluster is computed and further it looks for the user preferences in framing the clustered data unit in our proposed system. The advanced automated collaborative filter computes the rating similarity using Pearson Correlation Coefficient (PCC). newline