Utilizing Query by Example for Fast and Accurate Multimedia Retrieval

In this paper, several techniques have been introduced for improving the performance of Query-By-Example (QBE). First, a new crawler technique has been introduced to collect multimedia descriptive information. Second, a clustering technique is used for creating and updating multimedia databases. In addition, a new re-ranking algorithm which depends on the contents of the multimedia files and some user specific actions has been adapted. Furthermore, parallel technique has been used to decrease the query process time. Finally, elimination of duplicated files in multimedia query results has also been done. Several experiments have been studied and it has been observed that the proposed techniques allow the QBE to supports all multimedia types, increase the percent of relevant files and decrease the process time in query process