The Problems with Nomeclature and Crime Classifications

http://www.popvssoda.com/
Is it Soda or is it Pop? Is it a Theft or a Burglary?
You’d describe both objects the same, but depending on where you live in the country, you will call the same object by a different name. Soda vs Pop; Gumband vs Rubberband; Dad vs Pop; Clicker vs Remote vs TV Changer; Ya’ll vs Yins vs Yous; Pocketbook vs Purse vs Handbag; Monkeyball vs Gumball;  feels like the list is never ending!

This can also be applied to words when describing crimes on a crime map.

In categorizing thefts and burglaries, this word conundrum also happens. For example, cars.  When someone steals something from a car - is it a theft or burglary? For SpotCrime, we usually rely on the agency to classify these. Most Universities classify them as burglaries and many police agencies use theft.  

Shootings, however, are a different story for us. Most often a shooting is lumped into a category of aggravated assault. We break out shooting when we can - because our maps are designed for a public across the US. If the description the department lists that shots were fired or shooting, we will map the incident as a shooting. After all, wouldn’t you feel safer knowing where shootings occur compared to fist fights?

No classification is perfect, but with the current computing power and the power of search algorithms much of these classification problems can be solved with good descriptions. We believe there should be a bottom up approach taken instead of a top down approach.

Deciding specific crime classifications is an incredibly complicated and time consuming top down approach. Hundreds of classifications also put a burden on the officers in terms of keying the information. We think the emphasis should be on providing as much detail of the incident as possible - preferably using  as many descriptive words as possible - and as long as the descriptions are captured and indexed digitally, the departments can have computers scan and provide query results from the descriptions. This leads to a bottom up approach - similarly to how Google indexes the Internet and finds related websites when you type a couple of words into their search bar, we can have a similar index organizing crime data.

And for the record - it’s soda, not pop :)

Comments

Popular posts from this blog

SpotCrime Weekly Reads: Transparency, gun violence, crime data

SpotCrime Weekly Reads: AI, police conduct, transparency

SpotCrime Weekly Reads: Violent crime, AI tech, transparency