Tipping Points, Butterflies, and Black Swans: A Vision for Spatio-temporal Data Mining Analysis

摘要

Tipping points represent significant shifts that change the general understanding or belief of a given study area. The recent late winter 2011 events in the Mid-East and climate-level changes raise issues of whether such events are the result of random factors, tipping points, chaos theory or completely unpredicted black swans. Our vision is to understand how spatio-temporal data mining analysis can discover key variables and relationships involved in spatial temporal events and better detect when mining may give completely spurious results. One of the main challenges in discovering tipping point-like events is that the general assumptions inherent in any technique may become violated after an event occurs. In this paper, we explore our vision and relevant challenges to discover tipping point-like events in spatio-temporal environments.