Artisanal vs. Industrial: A look at approaches to content moderation

In a week when you might be discussing turkey preferences (free-range heirloom vs. supermarket Butterball), this new Data & Society paper feels appropriate: Robyn Caplan takes a look at approaches to content moderation, examining how 10 different platforms — from Facebook to Vimeo — handle a flow of user content. Caplan breaks the policies into three categories (which have also been discussed by Tarleton Gillespie in his book Custodians of the Internet.) She spoke with representatives from the companies about their approaches and also threaded in remarks that company executives made at the Content Moderation and Removal at Scale conferences earlier this year. and Here are some notes about each model.

Artisanal

— “Content moderation is done manually in-house by employees with limited use of automated technologies. Content moderation policy development and enforcement tend to happen in the same space. Platforms such as Vimeo, Medium, Patreon, or Discord Continue reading "Artisanal vs. Industrial: A look at approaches to content moderation"

What is the restaurant critic’s “duty to warn” in the age of the best-of listicle?

“I often think about our duty to warn. Mostly because, when we’ve neglected to, it’s ended up being a shitshow.” Thrillist has a fascinating piece from James Beard Award-winning food writer Kevin Alexander about how his rating of a small Portland, Oregon restaurant, Stanich’s, as Thrillist’s best burger in America inadvertently contributed to the restaurant’s closure five months later, as Stanich’s was overrun by one-off “burger tourists.” “Ever since we won this thing, people were going on the internet telling me how to cook a burger, telling me how screwed up we were, and frankly, that’s why I don’t go on the internet. It’s been the worst thing that’s ever happened to us. Tim McGraw came by and there was a five-hour wait and I couldn’t wait on him,” Steve Stanich told The Oregonian. Alexander felt absolutely horrible about this. After the Continue reading "What is the restaurant critic’s “duty to warn” in the age of the best-of listicle?"

Notifications every 2 minutes: This in-depth look at how people really use WhatsApp shows why fighting fake news there is so hard

This week, the BBC devoted attention to a series on the spread of false information: As part of its reporting, the BBC researched how ordinary citizens in India, Kenya, and Nigeria interact with fake news: “Participants gave the BBC extensive access to their phones over a seven-day period, allowing the researchers to examine the kinds of material they shared, whom they shared it with and how often.” This is really extensive research, including in-depth in-person interviews in multiple native languages across cities and regions plus the analysis of news in English and local languages across social
Continue reading "Notifications every 2 minutes: This in-depth look at how people really use WhatsApp shows why fighting fake news there is so hard"

Facebook probably didn’t want to be denying it paid people to create fake news this week, but here we are

Notifications every 2 minutes. The most interesting real-news-about-fake-news this week was the BBC’s in-depth research into how information — and misinformation — spreads via WhatApp in India. Read all about that here. (Also elsewhere on Nieman Lab, be sure to check out this piece by Francesco Marconi and Till Daldrup on how The Wall Street Journal is training its journalists to look out for deepfakes — the AI-generated videos that can make people appear to say and do things they really didn’t.) WhatsApp provides $1 million for misinformation research. Speaking of the globe-spanning chat app, which announced in July that it would fund misinformation-related research: WhatsApp said this week that it’s giving $50,000 each to 20 projects from 11 countries. Among the topics getting funding (Poynter’s Daniel Funke has the full list):
“Is correction fluid? How to make fact-checks on WhatsApp more effective” “Seeing is believing: Is video Continue reading "Facebook probably didn’t want to be denying it paid people to create fake news this week, but here we are"

Consumers love smart speakers. They don’t love news on smart speakers. (At least not yet.)

Smart speakers like Amazon Alexa and Google Assistant are rapidly gaining in popularity, but use of news on the devices is lagging, according to a report released Wednesday night by the Reuters Institute for the Study of Journalism. Use of the devices for music and weather is still far ahead of news use. And among consumers’ complaints about news briefings: They’re too long. Luckily, there’s time for news publishers to catch up, finds Nic Newman, a senior research associate at RISJ, who did his research via in-home interviews and focus groups, online surveys, and publisher interviews. (He also tapped Amazon, Apple, and Google for whatever data they were willing to share — which, unsurprisingly, wasn’t a lot; none of the companies would share data on how many devices they’ve sold or discuss trends in how news is consumed on them.) Smart speakers are still devices for early adopters:
Continue reading "Consumers love smart speakers. They don’t love news on smart speakers. (At least not yet.)"

How’s your newsletter doing compared to other news orgs’ newsletters? This tool lets you find out

If you’re a news organization sending out email newsletters with Mailchimp, you probably know that the company offers some metrics so you can track how your newsletters are doing. But if you want to know how your organizations’ newsletters are doing compared to those from other publishers, there hasn’t really been a way (beyond keeping your eyes open for random brags and gossip). And open rates can be a faulty metric. On Tuesday, the Shorenstein Center launched the “email benchmarking tool,” which news orgs can use both to track how their newsletters are doing according to six key metrics and to get a better picture of how their newsletters are doing to compared to others. (The tool was created as part of Shorenstein’s Single Subject News Project, which is funded by Knight; here’s some of its recent research on email newsletters.) Hong Qu, Shorenstein’s director Continue reading "How’s your newsletter doing compared to other news orgs’ newsletters? This tool lets you find out"

The New York Times is digitizing more than 5 million photos dating back to the 1800s

The New York Times is digitizing more than 5 million photos from its archives — some dating back to the 1800s — with help from a variety of Google technologies. The photos will be used in a series called Past Tense. (First up: a package focusing on how the paper covered California in the 20th century.) “Ultimately, this digitalization will equip Times journalists with useful tools to make it easier to tell even more visual stories,” Monica Drake, Times assistant managing editor, said in a statement. From CNET:
The newspaper’s “morgue” has 5 million to 7 million photos dating back to the 1870s, including prints and contact sheets showing all the shots on photographers’ rolls of film. The Times is using Google’s technology to convert it into something more useful than its current analog state occupying banks of filing cabinets. Specifically, it’s using Google AI tools to
Continue reading "The New York Times is digitizing more than 5 million photos dating back to the 1800s"