Sentiment Analysis

Analysis of TripAdvisor’s opinions for the tourist and cultural industry.

Since the advent of Web 2.0, the amount of social media data stored on the Internet has grown exponentially. Accordingly, public and private organizations are interested in analyzing the content of social media websites to study its reputation or know the opinion that is held regarding a product, a hotel, a restaurant, a political party or a particular cultural monument.

Sentiment Analysis is the task of Artificial Intelligence concerned to the analysis of the opinion of users, it thus suitable task for analyzing social media data.


Several works has been conducted focused on opinions of the tourism and cultural industry. TripAdvisor has been used as a data source. TripAdvisor is a travel Web 2.0, which shows opinions of travelers towards hotels, restaurants or tourist attractions about thousands of cities around the world. The knowledge of the opinion of the visitors is crucial for this type of institutions, because it is the best feedback to improve the quality of their services and the user experience. The outstanding results are:

  1. The viability of TripAdvisor as a source of opinions for cultural monuments was confirmed. The need to have a reliable source of data to develop our studies is crucial (see the paper https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=8012330).
  2. We proposed a methodology for the synthesis of opinions. An automatic summary of all negative opinions is made through the use of Deep learning and Subgroup Discovery methods. This can help the decision making processes of these organizations, detecting in a clear way those aspects that have to be improved (see the paper https://link.springer.com/article/10.1007/s12652-018-1150-3).

Period

January 2017 – Present

Resesarchers

Francisco Herrera, Ana Valdivia, Mª Victoria Luzón, Eugenio Martínez-Cámara.

Collaborator: Erik Cambria de Nanyang Technological University of Singapore.