Crime and security (8 articles)

GDP growth vs. criminal phenomena: data mining of Japan 1926–2013

Li, X., Joutsijoki, H., Laurikkala, J., & Juhola, M. (2018). AI & SOCIETY, 33(2), 261-274.

This article analyzes the correlation of economic progress and different criminal phenomena in Japan. Viscovery SOMine shows that GDP growth alone is not a good indicator to explain changes in crime rates. Economic stability seems to be more important than growth.

Crime vs. demographic factors revisited: application of data mining methods

Li, X., Joutsijoki, H., Laurikkala, J., & Juhola, M. (2016).

Criminal statistics and demographic constitution of 56 countries are compared. Viscovery SOMine is used for clustering and several machine learning approaches are used to confirm the validity of obtained clusters.

Homicide and its social context: analysis using the self-organizing map

Li, X., Joutsijoki, H., Laurikkala, J., Siermala, M., & Juhola, M. (2015). Applied Artificial Intelligence, 29(4), 382-401.

Homicide rates from 181 countries are analyzed with respect to socioeconomic driving factors. Viscovery SOMine is used to cluster this data using 62 of all 69 available attributes.

Application of data mining methods in the study of crime based on international data sources

Li, X. (2014).

In this dissertation, multiple crime data sets from all over the world are analyzed with respect to sociological, demographic and economic influences. Viscovery SOMine is used for clustering and visual data mining.

Application of the self-organising map to visualisation of and exploration into historical development of criminal phenomena in the USA, 1960–2007

Li, X., & Juhola, M. (2014). International Journal of Society Systems Science, 6(2), 120-142.

Crime and socioeconomic data in the United States from the years 1960–2007 are used to analyze the historical development and connections between crime and socioeconomic background. Viscovery SOMine is used to calculate a cluster model, which is also compared to results obtained by k-means clustering via different classification approaches.

Visualizing the influence of geography, oil and geopolitics on civil wars in the Arab world: a novel application of self-organizing maps and duration models

Mostafa, M. M., & Al-Hamdi, M. T. (2014). Civil Wars, 16(2), 239-254.

The influence of geopolitical factors, geography and resources on the duration and severity of civil wars in the Arab world is analyzed. Viscovery SOMine is used to analyze these influences visually and calculate clusters.

SOM neural network — a piece of intelligence in disaster management

Klement, P., & Snášel, V. (2009, December). In Nature & Biologically Inspired Computing, 2009. NaBIC 2009. World Congress on Nature and Biologically Inspired Computing. (pp. 872-877). IEEE.

Telephone calls to the European emergency number 112 are analyzed to detect anomalies and balance system load. Viscovery SOMine is used to cluster the calls according to incident type, district, time and other interesting variables.

Modular architecting for effects-based operations

Meteoglu, E. (2007).

This master thesis analyzes the effects of various diplomatic, military and other actions to further a strategic goal. Viscovery SOMine is used to cluster these actions according to multiple outcome dimensions.