A Machine Learning Method for Recognizing Invasive Content in Memes

In the time of web, Memes have become probably the sultriest subject on the web and apparently, the most widely recognized sort of satire seen via web-based networking media stages these days. Memes are visual outlines consolidated along with content which for the most part pass on amusing importance. Individuals use images to communicate via web-based networking media stage by posting them. Be that as it may, in spite of their enormous development, there isn't a lot of consideration towards image wistful investigation. We will likely foresee the supposition covered up in the image by the joined investigation of the visual and literary traits. We propose a multimodal AI structure for estimation investigation of images. According to this, another Memes Sentiment Classification (MSC) strategy is anticipated which characterizes the memes-based pictures for offensive substance in a programmed way. This technique uses AI structure on the Image dataset and Python language model to gain proficiency with the visual and literary element of the image and consolidate them together to make forecasts. To do such a process, a few calculations have been utilized here like Logistic Regression (LR), and so forth. In the wake of looking at all these classifiers, LR outbursts with an accuracy of 72.48% over the PlantVillage dataset. In future degrees, the use of labels related to online networking posts which are treated as the mark of the post while gathering the information.