Stock Trading Recommendations using Opinion Mining

Abstract

The prediction of stock trading factorsis critical in making decisions for investors. Most of them depend on disclosures of news for making decisions in selling or buying stocks. Nevertheless, precise stock market modelling trends through disclosures of news is a thought-provoking task, deliberating the ambiguity & intricacy of natural languages utilized. Unlike former work besides with the research that implements bag of words typically for extracting several features for building prediction method, we project a method based on opinions posted in twitter trends, which are highly influential towards a certain sector of the stock trading. A novel method that defines the meta-heuristics for trading factors listed as a quantity of stocks, opening cost, closing cost, aggregate cost and status. These meta-heuristic values of the corresponding factors recommend by their correlation with opinions exhibited in corresponding twitter trends. The experimental study was carried on the data extracted from the yahoo ticker, and the tweets posted during the contemporary period in twitter trends that are correlated to the yahoo ticks collected from the sector of the stock trades. The experimental study outcomes denoting the significance of the proposed stock trading recommendations using opinion mining (STROM).