MSUP: Model for the Searching of Unidentified People Comparing Binary Vectors Using Jaccard and Dice

摘要

The last years in Mexico were reported thousands of missing people. Almost every day were found peoples dead in some place. The authorities open a folder investigation, but most times is not possible an identification effective of the person. In other hand, the familiars report the missing of any member to Public Ministry (PM) or to National Search Commission (NSC). So, the authorities of PM or NSC document the personal data. However, there is not a connection between instances governmental. In this work, we propose a platform and algorithms to find missing people. These algorithms are: The first module is the Characterization of Unidentified People (CUP); the second module is the Characterization of Wanted People (CWP) and finally the Searching for Missing People (SMP). SMP will focus on an algorithm with similarity metrics with the ability to find one or more "similar" people found on the platform. This last module will determine which similar people were found, as well as the dependence on the government where they are physically. To implement the solution, it is necessary to: Establish a vector of characteristics obtained from the CUP module and the CWP module to apply algorithms based on similarity metrics until "matching" and evaluate the proposed algorithms to obtain the best result. For the solution we propose can benefit the institutions that have unidentified corpses under your responsibility. If a human remains is properly characterized, it increases the possibility of identifying and claiming. Therefore, it can reduce or avoid the problem of excess of thousands unclaimed corpses.