Sentiment analysis, an application of machine learning in business is the process of identifying and cataloging comments, reviews, tweets, feedback, and even random rants according to the tone or sentiments conveyed by it. The data is analysed using machine learning approach of Long Short Term Memory (LSTM) rating the sentiments on a scale ranging from −100 to 100. A new proposed activation function is used for LSTM giving best results as compared to the existing Artificial Neural Network (ANN) techniques. Depending upon the mined opinion, the business intelligence tools evaluate the products or services of a company eventually resulting in the increase of the sales of that company. The results clearly show that BI extracted from SA is quite instrumental in driving business effectiveness and innovation.

情感分析是机器学习在企业中的一种应用，它是根据注释，评论，推文，反馈甚至根据其传达的语气或情感随机识别和分类的过程。使用长短期记忆（LSTM）的机器学习方法对数据进行分析，以-100至100的等级对情感进行评级。与现有的人工神经网络相比，新提议的激活函数用于LSTM可提供最佳结果（ ANN）技术。根据获取的意见，商业智能工具会评估公司的产品或服务，最终导致该公司的销售额增加。结果清楚地表明，从SA提取的BI在推动业务效率和创新方面发挥了重要作用。