Web mining and minimization framework design on sentimental analysis for social tweets using machine learning

Web expansion includes various objective parameters such as storage devices, descriptive machines and third party optimization algorithms, making the web bulky and stacked for dynamic search results. Thus, web search optimization technique research provides various solutions. In this paper, a sentimental tweets segregation and classification based on content is proposed under the objective of web minimization for optimized search results. The methodology includes K Nearest Neighbouring (KNN) approach for extraction of similar tweets strength using the pre-learnt logs threshold values. Thus, appended to proposed novel framework designed for pattern extraction with respect to tweets logs and producing a results with 97.82% accuracy via real-time tweeter logs. The proposed framework is expanded towards pre-processed datasets of major social networking platforms to retrieve higher order of accuracy. The resultant outcomes are processed under reinforcement based machine learning for web minimization with an incurred enhancement of 1.75 refreshing rate with 4G LTE bandwidth. © 2019 The Authors. Published by Elsevier Ltd.