An Automated Framework For Dynamic Web Information Retrieval Using Deep Learning

The success of e-commerce through innovative web information retrieval technologies is crucial in today's world where customer feedback and information are highly sought after. Basic knowledge-gathering systems struggle to respond to dynamic changes on the webpage. This research uses the You Only Look Once (YOLO) algorithm and the Tesseract Long Short-Term Memory to develop a smart and flexible information extraction framework with convolutional and Long Short-Term Memory systems to facilitate automated dynamic Internet discovery. The Yolo Algorithm and Tesseract Long Short-Term Memory retrieve product descriptions found as images from Internet links. Experimental results include image recognition and 98% character recovery accuracy. Also, given the input database of 52 objects or images, an average of 75% accuracy is achieved.