Linguistic local change comparison of time series

In this paper a methodology is proposed for providing a linguistic summary of the comparison of time series. This series are obtained as a result of queries which the user defines over a data cube with time dimension. Our objective is that the final summary describe the similarity between series by using local changes. Definitions of the degree and sign of local changes are proposed. The understandability of the linguistic summary is achieved by means of linguistically quantified sentences. So are the coverage and brevity by using a hierarchical partition of time dimension inherited from the data warehouse. Our linguistic summaries are well suited to be included in an interface layer of a data warehouse system, improving the quality of human-machine interaction and the understandability of the results.