Linguistic database summarization is a capability needed by any intelligent system. Kacprzyk et al. achieved good results when they studied the problem of linguistic database summarization introduced by Yager (1982) and based on Zadeh's fuzzy theory. In this paper, the authors present a new method to generate linguistic summaries using hedge algebra structure. Compared with the fuzzy set approach, ours, based on hedge algebra, limits the loss of information and reduces bias in the data summary.