8: Discussion and Conclusion

Big data is a huge set of data which can be examined computationally to show patterns, associations and trends, particularly in link with human communication and behaviour. The revolution of big data occurs with the development of internet, smartphones, wireless networks, social media and other techniques. This chapter is based on the findings obtained from the analysis of data based on the reviews of TripAdvisor, in which customers used to share their experience. With the tourism sector, big data is one of the most essential concepts because several other firms are already using it and getting the recognition. This recognition involves the capability to make much informed decisions, study about rivalries and customers and develop the experience of customer and enhance revenue. Big data is used to align strategies to expectations of customers, adapt service proposals to new groups of customer and even create new opportunities of business. Big data is also used to anticipate where establishments can find a greater number of tourists and schedule their resources effectively. Some of the sectors have already started using machine learning and artificial intelligence techniques in search of delivering better quality service to their customers. And the tourism sector is not exempt from the exploration of betterment. The current study has also used machine learning techniques to classify the data. Data have been classified to figure out the preferences of customers, such as budget-friendly, cleanliness, ambience, service quality, etc. The tourism sector is a wide business area, but the keystone of this sector is the ambience part of it, and hotels are considered influential in that field. The complication of running a profitable business could be millstone and necessitate hotel organizations to hire more staffs which might dent their pocket to a large extent. Three different techniques have been used in the current study, i.e. association rule mining, pointwise mutual information (PMI) technique and neural network such as multiclass and multilabel classification algorithms. The study considered online hotel reviews and put forward a supervised machine learning technique employing unigram feature with different categories of information to carry out classification of documents. These three classification models have been used to check the accuracy of the results, i.e. TF-IDF, Doc2Vec and LDA. Various big data and machine learning techniques have been suggested and applied to tourism text data assessment for generating tourists' profile. After applying all the above-mentioned approaches, the study concluded that accuracy is high for the TF-IDF approach.