A New Multimedia Web-Data Mining Approach based on Equivalence Class Evaluation Pipelined to Feature Maps onto Planar Projection

Abstract

Multimedia information are semi-organized or unstructuredinformation elements whose essential substance is separately or by andlarge utilized for correspondence. Sight and sound information miningrecognizes, arranges, and recovers important highlights from an assort-ment of media to recognize enlightening examples furthermore, connec-tions for information acquisition. Computer Vision (CV)-based systemshave been increasingly popular in recent years, owing to the growingnumber and complexity of datasets. In CV, finding meaningful photosin a huge dataset is a difficult task to solve. Traditional search enginesretrieve photos based on text such as captions and metadata, but thisstrategy can result in a lot of irrelevant output, not to speak the time,effort, and money required to tag this textual data.In this paper, we proposed a pipelined deep learning oriented method-ology framework for multimedia web-data mining based on content ex-tracted feature maps in planner projection as input. Color, texture, form,and other high-level properties of images are represented as numericalfeature vectors. This technique is based on the following computer visiontasks in general i.e., Image segmentation, Image classification, Object de-tection etc. In order to prove the computational efficiency and to validateits statistical behaviour, we have also presented the experimental eval-uation on an standard multimedia dataset. The obtained performanceresults are then compared with some significant existing approaches inthe terms of various statistical measures/parameters.