Evaluation and distribution of text contents with models supported by artificial intelligence techniques in onion architecture

This thesis aims to analyze customer feedback using AI-supported models and distribute the obtained insights through onion architecture. The research specifically targets the airline transportation sector to enhance customer satisfaction and improve service quality by providing innovative methods. The thesis employs natural language processing (NLP) techniques and deep learning models to analyze text-based customer reviews. The first objective of the study is to extract meaningful insights from customer reviews by identifying topics and emotional responses through methods such as bigram, trigram analyses, and sentiment analysis. The second objective is to evaluate customer satisfaction through sentiment analysis, categorizing the reviews into positive, negative, or neutral sentiments. The third objective is to effectively distribute the analysis results using onion architecture. This architecture offers a layered and modular approach in software development, enhancing the sustainability and flexibility of applications. Lastly, the fourth objective is to develop decision support systems to improve service quality based on the data obtained from customer feedback. The dataset used in the research comprises text-based comments on airline experiences, including information such as verification status, review content, rating, recommendation status, and sentiment analysis. The dataset is pre-processed for sentiment analysis and split into training and testing sets. Using deep learning models and optimization algorithms, the results provide strategic recommendations for enhancing service quality and improving customer satisfaction.