Mining the News: Trends, Associations, and Deviations

News reports are an important source of information about society. Their analysis allows to understand its current interests and to measure the social importance of many events. In this paper, we use the analysis of news as a means to explore the society interests. We present a text mining technique that uncovers trends, discovers associations and detects deviations from news notes. The method uses simple statistical representations of the news reports (frequencies and probability distributions of topics) and statistical measures (the average or median, the standard deviation, and the correlation coefficient) for analysis and discovery of useful information. We illustrate the method with some results obtained from preliminary experiments and discuss their main implications.