An interdisciplinary analysis with data mining and visualization tools applied on multiple and multi-source time series - The case of the forest fund in Romania

Abstract

Starting from the idea of poor biodiversity conservation which continues to be one of the major both national and international current problems, this paper focuses on the analysis of the evolution of the forest fund and also on the validation of a set of hypothesis regarding specific indicators related to this evolution with a special consideration for interrelations and future trends based on a custom data mining model. The time series data was provided by the Romanian Ministry of the Environment, National Bank of Romania, Romanian Court of Accounts, National Institute of Statistics, Eurostat, Global Forest Watch, Global Forest Change, Google Earth and Transparency International. This data was filtered, calibrated, exported in order to be collected and integrated into a single online spreadsheet shared for open access and then progressively refined. It served as source of online representations under the form of interactive charts and maps or other visual metaphors able to capture the space-time evolution and distribution and provide basic support for querying. Finally, it was also the starting point for further time series analysis by using data mining tools. The results confirm the hypotheses formulated also on the basis of worrying satellite data and supported by the scientific literature. They also emphasize the influence of some historical moments associated with the appearance of specific laws progressively allowing an increasing private share of forests and also with the Romania’s alignment to the European Union on the evolution of some forest fund indicators.