IMPACT SCORE ESTIMATION WITH PRIVACY PRESERVATION IN INFORMATION RETRIEVAL

Abstract

Nowadays, Information Retrieval (IR) is becoming more popular technique due to the tremendous growth of resources on the internet. However, the present information retrieval techniques have several limitations such as lack of semantic keyword, more time consumption and vague user’s query, etc. To mitigate these issues, this paper proposed a novel Information Retrieval (IR) framework to achieve effective data access which is available in online. The proposed IR system includes five major steps, at first the documents which are shared as the resources are pre-processed, and domain analysis is made to find the category of the document. Secondly, the keywords are extracted using semantic keyword extraction and indexing, and impact score estimation is obtained to determine the importance of the keyword in each document. Thirdly, the document similarity is estimated using novel similarity estimation algorithm for clustering the documents based on the attained score. Fourth, the documents are ranked based on the similarity score and the impact score of the keywords in the query. Finally, the user needs to register their personal information based on the novel privacy preservation algorithm to maintain the privacy of the querying users. The simulation results of proposed framework achieved significant improvement than existing approaches in terms of average precision, recall, mean average precision and execution time.