Improving the Accuracy of Extracting Useful Information in Search Engines from the Web Using Deep Reinforcement Learning Based on the Q-Learning Algorithm

There is a huge amount of information, made accessible by exponential growth of the number of internet users and web content, which creates the challenge of efficient and accurate information retrieval. We propose a novel approach for improving search engine accuracy and speed by applying Q-learning algorithm. The proposed method optimises the retrieval of relevant web pages by leveraging of bloom filter and deep reinforcement learning with novel classification. The Jaccard criterion is used to evaluate the similarity of datasets of the web page characteristics and to classify and label them. This comparison measure maps the results into binary classes and then feed to Q-learning algorithm for further processing. Our technique improves search speed, memory consumption, and accuracy over existing techniques. This work offers insights into the theoretical development of large scale web data retrieval, which is a critical aspect of search engine optimisation and modern web information exploration.