What is it about?
This research explores how artificial intelligence can automatically determine whether online hotel reviews express positive or negative sentiments. The researchers collected over 40,000 English-language reviews from Booking.com and TripAdvisor about hotels in Tenerife, Spain. They tested different types of neural networks - specifically LSTM (Long Short-Term Memory) and CNN (Convolutional Neural Networks) models - to see which could best classify reviews as "good" or "bad" based on the text content. The study involved preparing the review data by removing punctuation and common words, then converting the text into numerical formats that the AI models could process. The researchers trained multiple versions of these models with different configurations and compared their accuracy in predicting sentiment. The LSTM models performed better overall, with the best model achieving 89.19% accuracy in correctly identifying positive and negative reviews.
Featured Image
Photo by Hugh Whyte on Unsplash
Why is it important?
This work addresses a critical need in the modern tourism industry, where online reviews heavily influence traveler decisions. With the explosion of user-generated content on platforms like TripAdvisor and Booking.com, tourism businesses need automated tools to quickly analyze thousands of reviews and understand customer sentiment at scale. The research has practical applications for hotel managers and tourism companies who can use these tools for competitive analysis, proactive customer service (responding quickly to negative reviews), reputation management, and risk assessment. As digital natives increasingly rely on peer reviews rather than corporate marketing content, having accurate sentiment analysis becomes essential for business success. The study's focus on Tenerife, a major tourist destination that received over 7 million visitors in 2016, demonstrates the real-world relevance of this technology for destinations dependent on tourism revenue.
Perspectives
Tourism represents the economic heart of our islands, and as a Canarian researcher, I have always felt a special responsibility toward this industry that sustains so many families in the archipelago. This project has given me a unique opportunity to merge my passion for artificial intelligence with something that has a direct and tangible impact on our community. What has motivated me most during this research has been the possibility of working closely with tourism industry professionals, from hoteliers to destination managers. Their practical perspectives have greatly enriched our technical approach, constantly reminding me that behind each analyzed review are real people's experiences and business decisions that affect the livelihoods of thousands of workers. Seeing how our models can help Tenerife hotels better understand their guests' opinions fills me with satisfaction. It's not just an academic exercise; it's a tool that can improve the tourist experience and, ultimately, strengthen our destination's competitiveness. I hope this work inspires more Canarian researchers to apply emerging technologies to local challenges, demonstrating that from the islands we can also contribute significantly to scientific and technological advancement.
Dr. Jesús Torres
Universidad de La Laguna
Read the Original
This page is a summary of: Using Deep Learning to Predict Sentiments: Case Study in Tourism, Complexity, October 2018, Hindawi Publishing Corporation,
DOI: 10.1155/2018/7408431.
You can read the full text:
Contributors
The following have contributed to this page
