International Journal on Future Revolution in Computer Science & Communication Engineering, 2017
The web contain huge amount of structured as well as unstructured data/information. This varying ... more The web contain huge amount of structured as well as unstructured data/information. This varying nature of data may yield a retrieval response that is expected to contain relevant response that is expected to contain relevant as well as irrelevant data while directing search. In order to filter out irrelevance in the search result, numerous methodologies have been used to extract more and more relevant search responses in retrieval. This work has adopted semantic search dealing directly with the knowledge base. The approach incorporates Query pattern evolution and semantic keyword matching with final detail to enhance significance of relevant data retrieval. The proposed method is implemented in open source computing tool environment and the result obtained thereof are compared with that of earlier used methodologies.
International Journal of Advanced Research in Computer Science, 2018
Nowadays, Information Retrieval (IR) is becoming more popular technique due to the tremendous gro... more 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.
International Journal on Future Revolution in Computer Science & Communication Engineering, 2017
The web contain huge amount of structured as well as unstructured data/information. This varying ... more The web contain huge amount of structured as well as unstructured data/information. This varying nature of data may yield a retrieval response that is expected to contain relevant response that is expected to contain relevant as well as irrelevant data while directing search. In order to filter out irrelevance in the search result, numerous methodologies have been used to extract more and more relevant search responses in retrieval. This work has adopted semantic search dealing directly with the knowledge base. The approach incorporates Query pattern evolution and semantic keyword matching with final detail to enhance significance of relevant data retrieval. The proposed method is implemented in open source computing tool environment and the result obtained thereof are compared with that of earlier used methodologies.
International Journal of Advanced Research in Computer Science, 2018
Nowadays, Information Retrieval (IR) is becoming more popular technique due to the tremendous gro... more 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.
Uploads
Papers by Kinjal Sheth