Algorithm Could Help Police with Gang-Related Crime Investigations

Gang crime is a very big problem in America; especially in large urban cities like Los Angeles.  UCLA and the LAPD have recently begun working together on a project that takes over a decade's worth of crime data, distinguishes regular crimes from perceived gang-related crimes, and then attempts to create an algorithm that identifies gang patterns and gang relationships.  This algorithm is also designed to identify which gangs were most likely involved in particular crimes.  In general, involvement by gangs is often hard to pinpoint - especially if trying to identify which gangs were involved because gang rivalries and alliances are constantly shifting and changing.  So far the algorithm can identify the three gangs most likely to be involved in a crime about 80% of the time and the one gang that was most likely involved about 50% of the time.  Even these odds are a lot better than the 17% chance that was produced without this algorithm by chance alone.  When attempting to solve crimes and combat gang-crime, even narrowing down the scope of what gangs might be involved helps because so many gangs are active in the Los Angeles area.  


What is being organized?

The things being organized here are the data and information on all crimes that police and police investigators in the Los Angeles area believe to be "gang-related" over the past 10 years.  This also includes any and all data on gang activity, culture, and related information.  This information is all organized chronologically before analyses are done.  


Why is it being organized?

The data is being organized for the purpose of trying to track and identify gang-related crimes in the Los Angeles area and to attempt to possibly predict the future of gang-related violence.  Patterns of gang-related data are constantly being searched out because gang-culture is extremely elusive and hard to understand.


How much is it being organized?

The data is being organized a lot.  I am sure that they have hundreds of categories that help describe how/what/when each crime occurred in order to get the best information and details about each crime instance.  Each "gang-related" crime instance is detailed and organized in the best possible way.  This type of organizing assists with being able to create the best possible algorithm to best understand gang-related crime.


When is it being organized?

The data is being organized as it is reported and logged into the system.  All crime data is logged as it happens in real-time.  However, I would imagine that this data was first compiled all at once as "gang-related" data at some period in the past 10 years and was then consistently updated as more crimes that were tagged as "gang-related" occurred.


By whom is it being organized?

This data is being organized by the Los Angeles Police Department (I'd imagine police, investigators, and analysts) as well as researchers at UCLA.  


http://articles.latimes.com/2011/oct/31/local/la-me-ucla-gang-computer-20111101