Anatomizing Text Without Training Dataset

Data is generated at a very fast pace every microsecond on various digital platforms such as micro-blogging channels, social media websites, product review applications, contentgenerating platforms, and many more. The Preponderance of this content generated on digital platforms is present in an unstructured format. In addition to the text, it also constitutes colloquial language, images, emoticons, audio, and even videos. So, it is of utmost significance to refine this unstructured data into the valuable form so that decisive insights can be brought from it. These insights can help businesses expand. Nowadays, companies invest heavily in text analytics applications such as recommender systems, customer grievance redress systems to enhance their sales and ultimately increase profit. Natural language processing and Text analytics have abetted organizations to yield informed and strategical decisions. Therefore, this field is gaining tremendous popularity. Many of these text analyticsbased applications like recommender systems require training data set to learn from the data and then produce a result for new data vectors. In this paper, we would like to propose an approach how text analytics can be performed without the need for training data set. It also follows the sequential steps for converting this unstructured data to a piece of structured information for getting high-valued insights. It also creates an analytics dashboard for visualizing data for easy interpretation.