A new model for linguistic summarization of heterogeneous data: an application to tourism web data sources

In this paper we present the problem of aggregating heterogeneous data from various websites with opinions about high end hotels into a database. We present the fuzzy model based on the semantic translation as a tool to obtain a linguistic summarization. The characteristics of this model (necessary to solve the problem) are not together on any of the existing linguistic models: the management of the input heterogeneous data (natural language included); the procurement of linguistic results with high precision and good interpretability; and the use of unbalanced linguistic term sets described by trapezoidal membership functions for defining the initial linguistic terms. We applied it to aggregate data from certain high end hotels websites and we show a case study using the high end hotels located in Granada (Spain) from such websites during a year. With this aggregated information, a data analyst can make several analyses with the benefit of easy linguistic interpretability and a high precision. The solution proposed here can be used to similar aggregation problems.