| Resumen |
Tourism in Mexico represents a primarily strategic activity for the country’s economy. Although the most renowned tourist spots generally have a wide promotion of their attractions, there are local businesses that are not sufficiently linked to this benefit and it is harder for the visitors to find them. Furthermore, social media is rich not only in reactions and opinions, but in experiences, these experiences can be useful to promote and recommend several sites. Web pages also have comments, similar to TripAdvisor, but small businesses are not present in it. In this sense, the techniques of Machine Learning can help to detect businesses with a good experience reflected in comments from web sites, in combination with data from social networks. Therefore, evaluating the quality of the tourist’s experience present in comments on social networks, and gathering personal information from apps represent an opportunity to generate not only recommendations, but itineraries based on time, space and experience. In this paper we present a framework to generate itineraries based on experiences of tourists in Mexico city, using machine learning and social mining, the results show similar performance for small-local business compared with the recommendations of popular and larger places. © 2020, Springer Nature Switzerland AG. |
