I can't see where the accuracy of such of a system would be very high. What kind of resources is a company going to deploy to verify that an employee is actually a risk after being flagged? What are they going to do after 6 months when 80% of their employees have been identified as risks?
Not when you get to cherry pick your data. If I throw out all the houses that had sold for low prices and pick only the 4 highest selling houses as the "comps", I skew the data to justify raising your taxes. In any housing market you will have homes on the market that have unreasonably high prices. Often someone will meet that asking price.
This brings up an interesting conversation I had with my brother the other day. He had recently received his city's tax assessment on his home in a major city in Canada. He felt that it was too high by about 25%. Looking at the list of supposed comparable homes in the area, none of them where really comparable. Different size, number of bedrooms/bathrooms, presence of attached garage, lot size, etc. One home was over a mile away in a completely different neighborhood. Several more comparable homes had recently (past 6 months) sold in the close neighborhood, but we're not listed as comps. The one thing they all had in common was that they had all sold for well above market value. So the question came up as to whether or not the algorithm used to find the comparable homes had been tuned to maximize (and justify) the tax revenue for the city. (My brother is appealing his assessment and is confident it will be lowered.)
I always thought that these type of incidents call for a "flash mob" type of response. Google "Joint Systems Manufacturing Center lima ohio" and click on Images and see top secret sensitive photos of the facility!
Which are "record emails"? All emails should be retained. So out the 1 billion emails in 2013, only 41,749 had to do with actual work? Seems we are paying a lot of people to not do any actual work.
Someone probably already has plans to incorporate the information into all the other marketing/marketable data on people and sell it.
What is the False Positive rate?
I can't see where the accuracy of such of a system would be very high. What kind of resources is a company going to deploy to verify that an employee is actually a risk after being flagged? What are they going to do after 6 months when 80% of their employees have been identified as risks?
No new toys...
No more new toys till you learn to take care of the ones you already have.
Re: Re: Computer algorithms/code used by local government should be published
Not when you get to cherry pick your data. If I throw out all the houses that had sold for low prices and pick only the 4 highest selling houses as the "comps", I skew the data to justify raising your taxes. In any housing market you will have homes on the market that have unreasonably high prices. Often someone will meet that asking price.
Computer algorithms/code used by local government should be published
This brings up an interesting conversation I had with my brother the other day.
He had recently received his city's tax assessment on his home in a major city in Canada. He felt that it was too high by about 25%. Looking at the list of supposed comparable homes in the area, none of them where really comparable. Different size, number of bedrooms/bathrooms, presence of attached garage, lot size, etc. One home was over a mile away in a completely different neighborhood. Several more comparable homes had recently (past 6 months) sold in the close neighborhood, but we're not listed as comps. The one thing they all had in common was that they had all sold for well above market value.
So the question came up as to whether or not the algorithm used to find the comparable homes had been tuned to maximize (and justify) the tax revenue for the city.
(My brother is appealing his assessment and is confident it will be lowered.)
Re: Throwing it open to all...
I always thought that these type of incidents call for a "flash mob" type of response.
Google "Joint Systems Manufacturing Center lima ohio" and click on Images and see top secret sensitive photos of the facility!
Why does the user/gov employee get to choose
Which are "record emails"? All emails should be retained. So out the 1 billion emails in 2013, only 41,749 had to do with actual work?
Seems we are paying a lot of people to not do any actual work.
Walkmart Logo
Perhaps they should go after the marketing team that came up with the Walmart logo.
It is very similar to the international symbol for asshole.