An analysis of opinions about tourist services based on the semantic orientation of texts

The analysis of opinions provided by international travellers is fundamental to promoting a successful Tourism Package (TP). In this study, we propose a methodology that can be used both with semantic orientation and with the classification of consumer opinions, that is, as either excellent or poor. The results from this classification can be provided to tourism operators with the aim of facilitating decision making with regard to what the clients want and say. In order to obtain the desired results, we utilized text mining algorithms and vectorial support machines - processes that enabled us to obtain acceptable level of precision and recall. In addition, it helped us obtain conclusions and gather data for new research projects in this area.