Connect with us

SOCIAL

Facebook Outlines New, Machine Learning Process to Improve the Accuracy of Community Standards Enforcement

Published

on

Facebook is always working to improve its detection and enforcement efforts in order to remove content that breaks its rules, and keep users safe from abuse, misinformation, scams, etc. 

And it’s systems have significantly improved in this respect – as explained by Facebook:

Online services have made great strides in leveraging machine-learned models to fight abuse at scale. For example, 99.5% of takedowns on fake Facebook accounts are proactively detected before users report them.”

But there are still significant limitations in its processes, mostly due to the finite capacity for human reviewers to assess and pass judgment on such instances. Machine learning tools can identify a growing number of issues, but human input is still required to confirm whether many of those identified cases are correct, because computer systems often miss the complex nuance of language.

But now, Facebook has a new system to assist in this respect:

CLARA (Confidence of Labels and Raters), is a system built and deployed at Facebook to estimate the uncertainty in human-generated decisions. […] CLARA is used at Facebook to obtain more accurate decisions overall while reducing operational resource use.” 

The system essentially augments human decision making by adding a machine learning layer on top of that which assesses each individual raters’ capacity to make the right call on content, based on their past accuracy.

Facebook CLARA system

The CLARA element is in the ‘Realtime Prediction Service’ sector of this flow chart, which assesses the result of each incident and crosschecks the human ruling against what the machine model would have predicted, while also referencing both against each reviewers’ past results for each type of report.   

That system, which has now been deployed at Facebook, has resulted in a significant improvement in efficiency, ensuring more accurate results in enforcement.

Compared to a random sampling baseline, CLARA provides a better trade-off curve, enabling an efficient usage of labeling resources. In a production deployment, we found that CLARA can save up to 20% of total reviews compared to majority vote.”

Which is important right now, because Facebook has been forced to reduce its human moderation capacity due to COVID-19 lockdowns in different regions. By improving its systems for accurately detecting violations, through automated means, Facebook is then able to concentrate its resources on the key areas of concern, maximizing the manpower that it has available.

Of course, there are still issues with Facebook’s systems. Just this week, reports emerged that Facebook is looking at a new way to use share velocity signals in order to better guide human moderation efforts, and stop misinformation, in particular, from reaching massive audiences on the platform. That comes after a recent COVID-19 conspiracy video was viewed some 20 million times on Facebook, despite violating the platform’s rules, before, finally, Facebook’s moderators moved to take it down.

Improved human moderation wouldn’t have helped in this respect, so there are still other areas of concern for Facebook to address, but the smarter it can be about utilizing the resources it has available, the more Facebook can focus its efforts onto the key areas of concern, in order to detect and remove violating content before it can reach big audiences.

You can read more about Facebook’s new CLARA reinforcement process here

Socialmediatoday.com

SOCIAL

Google Outlines Ongoing Efforts to Combat China-Based Influence Operations Targeting Social Apps

Published

on

Google Outlines Ongoing Efforts to Combat China-Based Influence Operations Targeting Social Apps

Over the past year, Google has repeatedly noted that a China-based group has been looking to use YouTube, in particular, to influence western audiences, by building various channels in the app, then seeding them with pro-China content.

There’s limited info available on the full origins or intentions of the group, but today, Google has published a new overview of its ongoing efforts to combat the initiative, called DRAGONBRIDGE.

As explained by Google:

In 2022, Google disrupted over 50,000 instances of DRAGONBRIDGE activity across YouTube, Blogger, and AdSense, reflecting our continued focus on this actor and success in scaling our detection efforts across Google products. We have terminated over 100,000 DRAGONBRIDGE accounts in the IO network’s lifetime.

As you can see in this chart, DRAGONBRIDGE is by far the most prolific source of coordinated information operations that Google has detected over the past year, while Google also notes that it’s been able to disrupt most of the project’s attempted influence, by snuffing out its content before it gets seen.

Dragonbridge

Worth noting the scale too – as Google notes, DRAGONBRIDGE has created more than 100,000 accounts, which includes tens of thousands of YouTube channels. Not individual videos, entire channels in the app, which is a huge amount of work, and content, that this group is producing.

That can’t be cheap, or easy to keep running. So they must be doing it for a reason.

The broader implication, which has been noted by various other publications and analysts, is that DRAGONBRIDGE is potentially being supported by the Chinese Government, as part of a broader effort to influence foreign policy approaches via social media apps. 

Which, at this kind of scale, is a concern, while DRAGONBRIDGE has also targeted Facebook and Twitter as well, at different times, and it could be that their efforts on those platforms are also reaching similar activity levels, and may not have been detected as yet.

Which then also relates to TikTok, a Chinese-owned app that now has massive influence over younger audiences in western nations. If programs like this are already in effect, it stands to reason that TikTok is also likely a key candidate for boosting the same, which remains a key concern among regulators and officials in many nations.

The US Government is reportedly weighing a full TikTok ban, and if that happens, you can bet that many other nations will follow suit. Many government organizations are also banning TikTok on official devices, based on advice from security experts, and with programs like DRAGONBRIDGE also running, it does seem like Chinese-based groups are actively operating influence and manipulation programs in foreign nations.

Which seems like a significant issue, and while Google is seemingly catching most of these channels before they have an impact, it also seems likely that this is only one element of a larger push.

Hopefully, through collective action, the impact of such can be limited – but for TikTok, which still reports to Chinese ownership, it’s another element that could raise further questions and scrutiny.

Source link

Continue Reading

SOCIAL

The Drum | Trump’s Instagram & Facebook Reinstatement Won’t Cause Marketers To Riot Yet, Experts Say

Published

on

The Drum | Trump's Instagram & Facebook Reinstatement Won’t Cause Marketers To Riot Yet, Experts Say

While the reinstatement of Donald Trump’s Twitter account in November had some advertisers packing up in protest, many will strike a different tune with Meta-owned Facebook and Instagram, experts predict.

Meta Wednesday announced that it’s lifting the ban on a handful of Facebook and Instagram accounts, including that of former US president Donald Trump – who was suspended nearly two years ago following the January 6, 2021 riots at the Capitol.

In a blog post yesterday, Nick Clegg, Meta’s president of global affairs, explained the reasons for the company’s decision, saying that it “evaluated the current environment” as it pertains to the socio-political landscape and security concerns and determined that “risk has sufficiently receded.” As a result, the company will welcome Trump back onto Facebook and Instagram.

The former president will be expected to comply with Meta’s user policies, but, considering his past violations, will face “heightened penalties for repeat offenses,” Clegg explained.

While it’s unclear whether Trump will become an active user on either platform following the decision, media and marketing experts are already sounding alarm bells at his potential return.

In particular, experts are cautious considering recent developments at Twitter. Elon Musk’s turbulent takeover – which has included mass layoffs, dramatic platform changes and the decision to reinstate the accounts of controversial figures like Trump and Kanye West (whose account has since been re-suspended) – has led to an exodus of advertisers. Could Meta’s decision to reintroduce Trump invite a similar fate?

‘Fear, frustration and protest’ could catalyze drawback

Concerns regarding brand safety and suitability on Facebook and Instagram are piquing among marketers. Trump’s presence on social media has long proven to exacerbate the spread of misinformation online. The risks of a potential recession, paired with new political tensions spurred by the 2022 midterms and the anticipation of the 2024 presidential election, may only up the ante.

“Misinformation on Meta’s platforms was an issue prior to Trump’s ban, during the ban and will likely continue to be an issue, even with the new [policies that] Meta has put in place,” says Laura Ries, group director of media and connections at IPG-owned ad agency R/GA. In light of this fact, Ries says, “Advertisers will need to continue to consider the type of content they’ll show up next to when evaluating whether or not to advertise on the platforms, especially as we march toward the 2024 election.”

She predicts that Meta may see some advertisers leave Facebook and Instagram “out of fear, frustration or protest.”

Others agree. “I suspect advertisers will not be pleased with this move and might make reductions in spend as they have done with Twitter,” says Tim Lim, a political strategist, PR consultant and partner at creative agency The Hooligans.

Although some advertisers are sure to pull back or cut their investments, the number will likely be low – largely because the scale and reach promised by both Facebook and Instagram will make it hard for most advertisers to quit. Smaller brands and startups in particular often rely heavily on Meta’s advertising business to spur growth, says Ries.

A ripple, not a wave

Most industry leaders believe Trump’s reinstatement won’t cause anything more than a ripple in the advertising industry. “Marketers who advertise on Facebook and Instagram care about their own problems, which generally [entail] selling more products and services,” says Joe Pulizzi, an entrepreneur, podcaster and author of various marketing books. “If Meta helps them do that, they don’t care one bit about brand safety – unless this blows up into a big political issue again. It might not, so marketers won’t do a thing.”

The sentiment is underscored by Dr Karen Freberg, a professor of strategic communications at University of Louisville, who says: “Facebook and Instagram are key fundamental platforms for advertisers. Marketers may … be aware of the news, but I am not sure if it will make a drastic change for the industry.” She points out that Twitter’s decision to lift the ban on Trump’s account in November caused such a big stir among marketers advertisers that Meta’s decision to do the same may come as less of a shock.

Trump’s return may even benefit Meta’s ads business by giving the company new opportunities to serve ads to Trump devotees, says Pulizzi. Ultimately, he says, Meta “needs personalities like Trump,” who, whether through love or hate, inspire higher engagement. “With Facebook plateauing and Instagram now chasing – and copying – TikTok at every turn, Trump’s follower base is important to Meta, which is hard to believe, but I think it’s true.”

But while some users may be energized by the former president’s return to Meta platforms, others may be outraged – even to the point of quitting Facebook and Instagram, points out Ries. In this case, she says, “advertisers will need to follow them to TikTok, Snap or other platforms where they’re spending their newfound time.”

R/GA, for its part, which services major brands including Google, Samsung, Verizon and Slack, will work on “a client by client basis” to address concerns about Facebook, Instagram or any other platform, says Ries. “R/GA recommended pausing activity on Facebook and Instagram after the insurrection and won’t hesitate to do so again if another incident occurs.”

For more, sign up for The Drum’s daily US newsletter here.

Source link

Continue Reading

SOCIAL

Snap Launches New Ad Campaign to Showcase its AR Offerings

Published

on

Snap Launches New Ad Campaign to Showcase its AR Offerings

Snapchat has launched a new promotional campaign which leans into the uniqueness of its viral AR trends, with a showcase of bizarre effects, as a means to present people with a different perspective on the real world.

Pretty trippy, huh?

As explained by Snap:

At Snap, we celebrate the joy, irreverence, and spontaneity of communicating with your real friends in fun, unexpected ways. Over the years, we’ve pushed the boundaries of how people see and experience the world through augmented reality. AR makes conversations and experiences better, and unlocks new ways to connect with others, learn about the world, shop, and more. [Our new campaign] shows you what it’s like to see the world the way Snapchatters do.”

It’s pretty weird, but will that get more people using Snap?

Certainly, the campaign will grab attention, and with 72% of active Snapchat users already engaging with AR elements in the app every day, there’s clearly a lot of interest in these types of weirdo activations that provide a new way of seeing the familiar.

Maybe that’ll prove to be a good lure to get people into the app, and broaden its user base. I mean, at the least, it’ll spark intrigue, which will likely get at least a few more people downloading the app to see what they can do.

AR is a key focus for Snap, and despite operating at a much smaller scale than Meta and Apple, which are both also investing big in AR projects, Snap has continued to punch above its wait in this area, by continually coming out with AR content that grabs attention, and engages audiences.

Meta is still struggling to maintain relevance with younger audiences, a key element that could de-rail its metaverse vision, while Apple has actually leaned on Snap to help showcase its advanced AR tools over time.

If nothing else, Snapchat has its finger on the pulse, which is why virtually every AR trend – from anime filters to baby faces, from crying faces to vomiting rainbows – all of these have originated from Snapchat, and that’s remained consistent over time, even with newer platforms like TikTok entering the same realm.

Snap is very in-tune with its user base, which is also why its Snapchat+ subscription offering is already doing better than Twitter Blue, even with the addition of tweet editing verification ticks (Snapchat+ has over 1.5 million paying subscribers, versus an estimated 325k for Twitter Blue).

That community sense has helped Snap maintain growth and relevance. But it also needs to expand – and maybe, through a bizarre showcase like this, that could help to make more people aware of the things that they can do in the app.

And this is how Snapchat Lenses tend to be shared. Somebody uses it, then they just have to show their friends.

In this respect, it seems like a good initiative, which could help Snap spark more interest and engagement.

It also serves as a demo of scanning in the Snap camera – if you want to try out any of the Lenses featured in the ad, you can scan the screen in the Snap camera, which will then open up whichever Lens is featured at that moment.

Source link

Continue Reading

Trending

en_USEnglish