Intelligent Visual Technique for an Assessment of Tweet Data Towards Social-Recommended Healthcare Solutions

Twitter has grown to be a vital social media platform for sharing healthcare knowledge, with over 300 million monthly active users. This paper addresses healthcare social recommendations using health tweets on social networks. Twitter's health-related tweet categorization is mainly dependent on topic models, which, unlike TF-IDF) (referred to as term frequency and inverse document frequency), discover topics (or health clusters) inside unlabeled tweets. Traditional topic models are used to extract the characteristics of tweets and model those characteristics. Visual assessment of clustering tendency (VAT) and cosine-based VAT (cVAT) are two state-of-the-art visual techniques for analyzing health data clusters of tweets. In the proposed work, intelligent multiple perspective cosine similarity-based VAT (IMPCS-VAT) is developed, which has achieved remarkable success in finding health clusters regarding health issues and solutions using social health data. It considers multiple perspectives while calculating similarities while assessing the health tweets for the social recommended healthcare solutions. Reliable findings from clustering health tweets require multiple perspectives to access more useful similarity features across tweet documents. The experimental study is well illustrated with visual health clusters by the proposed visual technique to determine health tweets' topics in an intelligent mechanism.