An algorithmic implementation of entropic ternary reduct soft sentiment set (ETRSSS) using soft computing technique on big data sentiment analysis (BDSA) for optimal selection of a decision based on real-time update in online reviews

In our day to day life, we face so many decision-making problems. The heavy text data sets and these data sets are increasing drastically in such a way that it reaches to the big data environment. Here, We have not only proposed a framework for big data sentiment analysis on real-time updates in online reviews or text for optimal or best decision selection (for example selection of a restaurant) from existing huge list of N number of restaurants but also implemented our proposed framework as a mathematical algorithm (named as Algorithm 4.1) by using soft computing technique for finding reduct soft set of consolidated review matrix. We further quantified sentiments in three values (1, −1, and 0), either in 1 for (positive/yes/true) or −1 for (negative/no/False) and 0 for (neutral or absence of sentiment) and stored them in a table (named as ternary sentiment table). Then, we have done an entropic calculation on this ternary sentiment table to find the quantity of information stored in its associated rows and columns. This proposed quantification further helps to identify the most important attribute of the table. It helps to decide weight for the different attributes and applying calculated weights to corresponding attributes to obtain the quantified ordered decision-making values.