Discussion on the paper by Bernard W Silverman on "Multiple-systems analysis for the quantification of modern slavery: classical and Bayesian approaches"

Kuldeep Kumar (Bond University, Gold Coast)

I join the chorus that there is a strong need for more research in estimating the victims of heinous crimes such as human trafficking, especially for laying down the foundation for policy implementation. The present study reviews various methods and investigates their performance on a range of real data sets. My suggestion to Professor Silverman would be to utilize the rich data set that he has for various other research avenues using graph analysis techniques which include link prediction analysis, clustering and visualization. These techniques can be used to find suspicious individuals, existing network relationships, unusual changes and anomalous network structure as well as geospatial dispersions. Besides, the visualization of dynamic and complex data in the form of graphs leads users to have a detailed overview of the data, to filter, select and look into networks details. I would be personally keen to have a look at the data and to use the expertise that is at my disposal to do the same and to take advantage of the hard work and commendable efforts put in by Professor Silverman in the collection and compilation of the data from several sources.