<![CDATA[This research aimed at improving the reliability of the system by developing a machine learning framework as a relevance feedback mechanism that aims to harness human perception in the training system to learn to map between the vector feature extraction results to the specifications given query. In the developed relevance feedback mechanism, users are involved in providing feedback to the system via the query by example (QBE) interface model. Basically, this module works with the user to decide how to involve the relevant results or irrelevant that emerged as a result of the system output user-defined queries. Then the machine-learning mechanisms will reformulate return query results based on user ratings and displays the new results. This process can take place iteratively until the user is satisfied on the relevance of the results obtained. Outcomes of this study will produce an image search system that can be integrated with information systems of third party services through the modules developed]]>
Copyrights © 2016
