Automatic Generation of Restaurant Reviews Using Natural Language Processing

Currently, there is plenty of information available on restaurant rating platforms, but sometimes it can be contradictory or difficult to analyze in depth. An important challenge for consumers is searching for useful opinions and making decisions based on reviews usually obtained from different social networks or rating platforms. This project addresses this issue through the design, development, and implementation of a system that generates global recommendations for a set of restaurants based on the analysis of reviews using Natural Language Processing (NLP) techniques. The system is based on a corpus of restaurant reviews in Mexico City, from which relevant aspects are extracted and synthesized to generate comprehensive reviews from various sources. The system can also evaluate customer satisfaction by identifying the positive and negative aspects mentioned in their reviews. In this way, it provides comprehensive information that helps diners make informed decisions. By gathering data from various sources, the system classifies and analyzes the information, providing an analysis rather than just displaying data. Another important aspect is that the project contributes to the promotion of the gastronomic offer in Mexico City, supporting tourism in a more informed way. By integrating customer perspectives, a more complete and realistic view of the restaurant experience is obtained. The importance of this project lies in the empirical evidence showing that consumer reviews are influenced by the average rating and the number of reviews. Given the overwhelming number of options and the need to provide relevant information efficiently, this project offers a solution by generating detailed reviews based on aggregated information from multiple sources, including consumer reviews and influencer critiques.