SoylentNews is people
SoylentNews
from the Who-LiveStream's-the-Watchmen? dept.
The Root reports that “Almost half of Americans hate their police department:”
[DrugAbuse.com] examined over 766,000 tweets about sentiment toward law enforcement in each state. The state with the most positive perception of police was New Hampshire. The most negative: Arkansas. The city with the most positive perception of police was Columbus, Ohio, while the one with the most negative was, not surprisingly, Ferguson, Mo. Other “failing” city police departments included Los Angeles, Miami, Phoenix, New York and Denver. Baltimore, a city still reeling from recent unrest, received a D grade….
“If you talk to young people in Baltimore, I don’t think their feelings about police have changed at all in the last five to seven years,” says [Philip Leaf, a Johns Hopkins University professor]. “There has been a negative perception of police in many communities for a long time. There just haven’t been conversations with these young people or in the media about it until recently. There hasn’t been an upsurge of disconnect with the police. With cellphones, there has been documentation of things that people have been talking about for a long time. People haven’t been believed, and now it’s hard not to believe it, if you see it on TV.…”
“It’s not as if this stuff hasn’t been going on all along for decades, but now it’s being captured for the world to see, and the few bad apples being captured on camera are ruining the entire tree of law enforcement,” says Hassan Giordano, 39, and a candidate for Baltimore City Council. “However, those very same people who have a negative opinion of police will also be the same ones calling 911 when they find themselves in an unsafe situation. It’s a catch-22.”
It's important to note that on the graphs shown in the article, even an A grade represents negative sentiment.
More data and a description of the methodology are available at DrugAbuse.com, including graphs of tweet sentiment involving alcohol, drugs, and marijuana. DrugAbuse.com used the commercial IBM service AlchemyAPI to analyze the tweets.
(Score: 2, Insightful) by SunTzuWarmaster on Wednesday August 26 2015, @09:05PM
(should have used Preview)
It can be done right; an example might include relative rankings of airlines. No airline has good reviews (people that tweet about airlines aren't tweeting about hot stewards or tasty peanuts; airlines that leave on time and arrive on time don't get tweets), but, scaled and taken in summation, they can provide insight. Data like "1/100 people that take Delta are dissatisfied, but 20/100 that take JetBlue are dissatisfied" presents a meaningful message. This message may instead be presented, with an unequal traffic assumption, as "Jetblue has 40x more negative tweet-action than Delta, despite only having twice as many flights".
Another way it can be done right is to scale by traffic. "Of the people who are tweeting, 12% of them say they love Coke, 8% of them say they love Pepsi, and 9% of them say they love RC; Pepsi and Coke have similar brand penetration, at 43% and 40%, while RC has roughly 10% of the market." This paints a picture of a "small, but rapid, following of RC lovers".
That said, the authors didn't scale their per-city results by population, or by percentage of traffic. Presumably out of laziness*. Instead, they end up with a population heat map problem (https://xkcd.com/1138/).
*If you are an author and you did, you should report it. I read both linked articles.
Full Disclosure: My (one-person) company had a sentiment analysis product pitched/demo'd to potential investors, where it was not funded.
(Score: 3, Interesting) by VLM on Wednesday August 26 2015, @09:50PM
That said, the authors didn't scale their per-city results by population, or by percentage of traffic
I'd like to see it corrected for level of criminal activity. Fundamentally I think most high crime areas are always going to be unhappy with the cops, either because they're directly opposed to the cops, or the cops are so pissed off from dealing with the opposition that they shoot random people in frustration, leading to unhappiness.
For example the south side of Chicago is always going to be more unhappy with the cops, per thousand residents, than rural northern Wisconsin, for example.
(Score: 2) by Anal Pumpernickel on Thursday August 27 2015, @02:53AM
High crime areas are always going to be unhappy with the cops because they have to deal with the cops more often, and therefore know just how often cops unjustly abuse people's rights while the ones who don't do the abuses themselves defend or cover for the ones who do. It's not just about cops murdering people; there are many rights violations that do not end in death.