Analyzing Polls and News Headlines Using Business Intelligence Techniques

Opinion and market research companies gather a substantial amount of polls data which can be combined with news headlines, for the corresponding time periods they are collected. These data are analyzed in order to answer specific (predefined) questions related to the situation of each time period. However, when these tasks are fulfilled, the collected data are archived and possibly the majority of them will remain unutilized for future research. In this paper, we argue that these „inactive‟ data can be further analyzed and hidden knowledge can be extracted. For this reason, we propose an appropriate framework based on modern Business Intelligence (BI) techniques. The innovation of the proposed framework is that it is able to reuse and analyze data that have been collected in the past and discover hidden knowledge, which can be utilized to bring profit in many ways. The basic scope of our framework is a) to supply knowledge on trends regarding specific politico-social and market issues and the way in which the situation of each period affect those trends, b) to predict the evolvement of trends according to the current situation or the formulation of emerging trends.