A Cloud-based Machine Learning Pipeline for the Efficient Extraction of Insights from Customer Reviews

Abstract-- The efficiency of natural language processing the text may be considered noise, depending on how we view has improved dramatically with the advent of the data and what we think is relevant. This in turn makes machine learning models, particularly neural networkbased it difficult to solve this task: it is not enough to find some solutions. However, some tasks are still challenging, specific information in texts, but we also have to decide what especially when considering specific domains. In this information we need based on the texts. Our application is an end-to-end methods integrated into a pipeline. For topic modeling, system that runs on a cloud-based infrastructure and can be our composite model uses transformer-based neural networks used as a service by small and medium-sized businesses. Furthermore, our some aspects, outperforms currently available approaches. Our environment, we needed a model that could identify topics approach is validated and compared with other stateof-the-art in texts and provide a way to determine which topics are relevant, methods using publicly available datasets for given our analysis goals. We focused on these tools Users of social platforms, forums, and online stores generate because we found them to be the most appropriate for our a significant amount of textual data. One of the most goals based on the available literature. It was also important useful applications of machine learning-based text processing to us that these methods and complex models have a stable is to find words and phrases that describe the content of implementation, are verifiably executable in our cloud-based these texts.