Ranking System for All the Tourism Related Hotel Industries Using NLP and ML Approach

Globally, people witness a booming multi-dollar revenue generation from the hotel industry which increases consistently. This approach is planned to develop a system that ranks different hotel industries worldwide or have a localized ranking for a city or province etc., which considers peoples reviews as its base. Having sufficient number of customers reviews available online help us in this regard. The idea is to analyse the most appreciated tourism related hotel industry to understand the attributes contributing to the revenue. The dataset is acquired from Kaggle which is related to worldwide hotel industries reviews. A novel approach, that uses the ensemble of binary classification called Bidirectional Encoder Representations from Transformers (BERT) model to learn the contextual relations between the words that are necessary for sentence embedding, is implemented. Later, the weighted averages take into account the relative importance and frequency of some factors in our dataset. Using these values, this approach will be able to rank different tourism related hotels. This system can further be enhanced by different stakeholders to plan future ventures based on the results from our model.