e dot dot dot
a mostly about the Internet blog by

October 2020
Sun Mon Tue Wed Thu Fri Sat
       


Content Moderation Case Study: Suppressing Content To Try To Stop Bullying (2019)

Furnished content.


Summary: TikTok, like many social apps that are mainly used by a younger generation, has long faced issues around how to deal with bullying done via the platform. According to leaked documents revealed by the German site Netzpolitik, one way that the site chose to deal with the problem was through content suppression -- but specifically by suppressing the content of those the company felt were more prone to being victims of bullying.The internal documents showed different ways in which the short video content that TikTok is famous for would be rated for visibility. This could include content that was chosen to be featured (i.e., seen by more people) but also content that was deemed Auto R for a form of suppression. Content rated as such was excluded from the for you feed on Tiktok after reaching a certain number of views. The for you feed is how most people view TikTok videos, so this rating would effectively put a cap on views. The end result was the reach of content categorized as Auto R was significantly limited, and completely prevented from going viral and amassing a large audience or following.What was somewhat surprising was that TikTok's policies explicitly suggested putting those who might be bullied in the Auto R category -- even saying that those who were disabled, autistic, or with Down Syndrome, should be put in this category to minimize bullying.

According to Netzpolitik, employees at TikTok repeatedly pointed out the problematic nature of this decision, and how it was discriminatory itself and punishing people not for any bad behavior, but because of the belief that their differences might possibly lead to them being bullied. However, they claimed that they were prevented from changing the policies by TikTok's corporate parent, ByteDance, which dictated the company's content moderation policies.Decisions to be made by TikTok:
  • What are the best ways to deal with and prevent bullying done on the platform?
  • What are the real world impacts of suppressing the viral reach of any content based on the type of person making the content?
  • Is it appropriate to effectively prevent those you think will be bullied from getting full access to your platform to prevent the possibility of bullying?
  • What data points are being assessed to justify the assumptions being made about Auto R being an effective anti-bullying tool?
Questions and policy implications to consider:
  • When there are strong pushes from policymakers to platforms that they need to stop bullying will it lead to unintended consequences like the effective minimization of access to these platforms by potential victims of bullying, rather than dealing with the root causes of bullying?
  • Will efforts to prevent a bad behavior be used to really sweep that activity under the rug, rather than looking at how to actually make a platform safer?
  • What is the role of technology intermediaries in preventing bad behavior?
Resolution: TikTok admitted that these rules were a blunt instrument that were put in place rapidly to try to minimize bullying on the platform -- but that the company had realized it was the wrong approach and had implemented more nuanced policies:
"Early on, in response to an increase in bullying on the app, we implemented a blunt and temporary policy," he told the BBC."This was never designed to be a long-term solution, and while the intention was good, it became clear that the approach was wrong."We have long since removed the policy in favour of more nuanced anti-bullying policies."
However, the Netzpolitik report suggested that this policy had been in place at least until September of 2019, just three months before its reporting came out in December of 2019. It is unclear exactly when the more nuanced anti-bullying policies were put in place, but it is possible that they came about due to the public exposure and pressure from the reporting on this issue.

Read more here

posted at: 12:00am on 08-Oct-2020
path: /Policy | permalink | edit (requires password)

0 comments, click here to add the first



16K COVID-19 Cases Go Missing In UK Due To Government's Use Of Excel CSVs For Tracking

Furnished content.


Yes, yes, you're sick of hearing about COVID-19. Me too. But the dominant force of 2020 continues to provide news, often times with a technology focus. This mismanaged pandemic has already given us an explosion of esports, students gaming remote learning systems, and enough dystopia to make George Orwell vomit in his grave.But to really get your anger bubbles gurgling, you need turn only to the myriad of ways far too many governments have taken a moment that requires real leadership and forethought, and pissed it all down their legs. America appears to be trying to lead the charge in this, with our shining city on the hill mostly being illuminated by headlights of cars carrying sick passengers looking to get tested for this disease. Still, we're not alone when it comes to sheer asshatery. The UK recently managed to lose thousands of COVID-19 cases... because it was tracking them in Excel CSVs.

The issue was caused by the way the agency brought together logs produced by commercial firms paid to analyse swab tests of the public, to discover who has the virus. They filed their results in the form of text-based lists - known as CSV files - without issue.PHE had set up an automatic process to pull this data together into Excel templates so that it could then be uploaded to a central system and made available to the NHS Test and Trace team, as well as other government computer dashboards.
Public Health England (PHE) decided to put all of this information into a file using the XLS format. XLS was first introduced in 1987 and was replaced by the XLSX format over a decade ago. Putting aside the use of Excel to monitor positive COVID-19 cases in a major industrialized nation for just a moment, just the use of an antiquated format managed to lose PHE over sixteen thousand positive cases.How? Well, XLS has restrictions as to how many rows of data it can record.
As a consequence, each template could handle only about 65,000 rows of data rather than the one million-plus rows that Excel is actually capable of. And since each test result created several rows of data, in practice it meant that each template was limited to about 1,400 cases.When that total was reached, further cases were simply left off.
Which means the people that had COVID-19 weren't tracked for contact tracing. The government and its people didn't have a complete picture as to either the total case count for the disease, nor its positivity rate. In other words, the agency in charge of national health failed to keep the nation informed as to its risk exposure because it didn't know how to properly use a common office application that it repurposed to record COVID-19 data.
Labour's shadow health secretary Jonathan Ashworth said lives had still been put at risk because the contact-tracing process had been delayed."Thousands of people [were] blissfully unaware they've been exposed to Covid, potentially spreading this deadly virus at a time when hospital admissions are increasing," he told the House of Commons. "This isn't just a shambles. It's so much worse."
The UK's Health Secretary told the House of Commons that PHE had decided to replace the use of Excel, or what he called a "legacy system", two months ago. But apparently PHE hadn't gotten around to it yet.And still hasn't, actually. In fact, PHE's plan to temporarily fix all of this is... more Excel!
To handle the problem, PHE is now breaking down the test result data into smaller batches to create a larger number of Excel templates. That should ensure none hit their cap.But insiders acknowledge that the current clunky system needs to be replaced by something more advanced that excludes Excel, as soon as possible.
When you hear complaints that governments are not taking this pandemic seriously, this is what they mean.

Read more here

posted at: 12:00am on 08-Oct-2020
path: /Policy | permalink | edit (requires password)

0 comments, click here to add the first



October 2020
Sun Mon Tue Wed Thu Fri Sat
       







RSS (site)  RSS (path)

ATOM (site)  ATOM (path)

Categories
 - blog home

 - Announcements  (1)
 - Annoyances  (0)
 - Career_Advice  (0)
 - Domains  (0)
 - Downloads  (3)
 - Ecommerce  (0)
 - Fitness  (0)
 - Home_and_Garden  (0)
     - Cooking  (0)
     - Tools  (0)
 - Humor  (1)
 - Notices  (0)
 - Observations  (1)
 - Oddities  (2)
 - Online_Marketing  (146)
     - Affiliates  (1)
     - Merchants  (1)
 - Policy  (2167)
 - Programming  (0)
     - Browsers  (1)
     - DHTML  (0)
     - Javascript  (5)
     - PHP  (0)
     - PayPal  (1)
     - Perl  (37)
          - blosxom  (0)
     - Unidata_Universe  (22)
 - Random_Advice  (1)
 - Reading  (0)
     - Books  (0)
     - Ebooks  (1)
     - Magazines  (0)
     - Online_Articles  (4)
 - Resume_or_CV  (1)
 - Reviews  (1)
 - Rhode_Island_USA  (0)
     - Providence  (1)
 - Shop  (0)
 - Sports  (0)
     - Football  (0)
          - Cowboys  (0)
          - Patriots  (0)
     - Futbol  (1)
          - The_Rest  (0)
          - USA  (1)
 - Windows  (1)
 - Woodworking  (0)


Archives
 -2020  December  (8)
 -2020  November  (46)
 -2020  October  (48)
 -2020  September  (49)
 -2020  August  (47)
 -2020  July  (46)
 -2020  June  (46)
 -2020  May  (49)
 -2020  April  (48)
 -2020  March  (47)
 -2020  February  (46)
 -2020  January  (48)


My Sites

 - Millennium3Publishing.com

 - SponsorWorks.net

 - ListBug.com

 - TextEx.net

 - FindAdsHere.com

 - VisitLater.com