Big Textual Data Analytics Using Transformer-Based Deep Learning for Decision Making

With the remarkable emergence of significant results for various deep learning-based transformation techniques, which are pre-training models such as BERT and its branches, which have the potential to deal with big data analysis frameworks. This represents a qualitative leap in understanding large-sized textual data of the opinions of Web users in order to classify it into several poles, and it is an important motivation for valuing it and benefiting from it in Decision-Making by managers of various companies such as marketing, health care insurance, finance, protection, etc. This paper explains how to provide high performance in analyzing huge text data by building a solid model suitable for analyzing huge text data, then classifying it by improving the contextual information in the sentence using the BERT technique with mechanism CNN. Extensive experiments on large-scale text data have demonstrated the remarkable efficiency of our model, an estimated percentage 92% compared to new and recent research studies.