Applied Machine Learning Techniques for Classification Reviews of Insurance Company

Sentiment analysis of the texts is one of the most significant fields in the science of computer linguistics and natural language processing (NLP). Sentiment analysis is used in websites with reviews, recommendation systems, business analysis, etc. The goal of the work is to research and to apply machine learning techniques for sentiment analysis of insurance company reviews. The concept of sentiment, approaches to determining the sentiment, methods of text preprocessing were studied during the research. Clustering and classification methods for unstructured big data were studied and implemented in this work. The data for the study were extracted from an open source with customer reviews. A model for determining the sentiment of texts has been created on the Python language. The research results are an overview of how machine learning methods work, a text sentiment model and analyzed reviews. Analytical hypotheses were put forward on the basis of classified data. The results could be used by analysts to set customer preferences and solve business problems. The work contains 60 pages, includes 22 images and 4 tables. In the list of references 28 sources were used.