Federated Learning

When I first think of federated learning, what comes to mind is something like a college federated department. For example, the history faculty at NJIT and Rutgers University-Newark are joined in a single federated department offering an integrated curriculum and joint undergraduate and graduate degree programs.

Having worked at NJIT, it made sense to combine the two departments and collaborate. Each had its own specialties but they were stronger together.

In technology, a federation is a group of computing or network providers agreeing upon standards of operation in a collective fashion, such as two distinct, formally disconnected, telecommunications networks that may have different internal structures.

There is also federated learning which sounds like something those two history departments are doing, but it is not. This federated learning is the decentralized form of machine learning (ML).

In machine learning, data that is aggregated from several edge devices (like mobile phones, laptops, etc.) is brought together to a centralized server.  The main objective is to provide privacy-by-design because, in federated learning, a central server just coordinates with local clients to aggregate the model's updates without requiring the actual data (i.e., zero-touch).

I'm not going to go very deep here about things like the three categories (Horizontal federated learning, vertical federated learning, and federated transfer learning). As an example, consider federated learning at Google where it is used to improve models on devices without sending users' raw data to Google servers.

comic
An online comic from Google AI

For people using something like Google Assistant, privacy is a concern. Using federated learning to improve “Hey Google,” your voice and audio data stay private while Google Assistant uses it.

Federated learning trains an algorithm across the multiple decentralized edge devices (such as your phone) or servers that have local data samples, without exchanging them. Compare this to traditional centralized machine learning techniques where all the local datasets are uploaded to one server.

So, though federated learning is about training ML to be efficient, it is also about data privacy, data security, data access rights and access to heterogeneous data.


MORE at analyticsvidhya.com...federated-learning-a-beginners-guide
 

Fake Facebook Accounts

signupNo one is giving Facebook or Meta or Mark Zuckerberg a free pass these days. Criticism is a daily event and their PR people must be in constant firefighting mode. But Facebook has been doing things to protect privacy and security and Facebook I keep hearing ads on podcasts about how they are promoting safety.  Unfortunately, the criticism is usually drowning out the tools they do offer and the actions they are taking. If you are on Facebook, you should be doing your part in protecting your account. Much of online protection is a matter of personal responsibility.

One of the areas that often gets attention is fake accounts. By the end of 2019, Facebook removed a staggering 5.5 billion fake accounts. Plenty of companies would be happy to have that many legitimate accounts but Facebook far exceeds that number. The removals continue and from the fourth quarter of 2017 to the third quarter of 2021 they removed approximately 1.8 billion fake accounts, up from 1.3 billion fake accounts in the corresponding quarter in 2020.

Why would anyone want to make a fake account?

Scammers use fake Facebook accounts to connect with users and then their friends to scrape personal information. That can be used to steal identities. They can also reach out to anyone who's accepted that fake friend request to try and scam them, since this fake account is sending friend requests to all your friends. This is called Facebook cloning.

I have seen a number of my friends' accounts be cloned using a few photos they have made public and any "public" information. When a clone of your account is created using your name you are not the real object of attention. It is actually your friends that are the target and the hope is that your friends will accept the fake friend request.

The cloned account often looks quite bare. Some people immediately recognize that they are already friends with that person/name and know this new request is fake. But if you have many friends or accept requests without looking a bit closer, you can be scammed. If you accept that friend request, the scammer now has access to your friend-visible information and then your friends' accounts' personal information. This access expands rapidly even if not everyone accepts the requests.

To report a fake account, go to  https://www.facebook.com/help/306643639690823

You should also take the initiative to use some tools Facebook offers and do some protecting on your own.

Privacy Checkup - helps you control who can see what you share, how your information is used and how to secure your account.

You see the message if you click next to the ad "Why Am I Seeing This Ad?" You can adjust what ads you see starting with that one.

Off-Facebook Activity is a tool most users don't know about. It lets you control or disconnect the information businesses send to Meta about your activity on other apps and websites.

And you can download or request and export your data so you can move it between services.

Google and YouTube Changing Features for Kids and Teens

social media

Image by Gerd Altmann from Pixabay

With continuing pressure on the big tech companies to protect user privacy, particularly for younger users, Google is introducing updates to YouTube and its search feature aimed at increasing safety for kids and teens on its platforms. The changes include a number of things to give minors more control over their digital footprint and somewhat constrain commercial content for children.

Some of these changes affect their bottom line profits but there is the PR value of making these changes, and I'm sure they hope it will keep the government from regulating or punishing them for awhile.

Google stated that it wants to work with "kids and teens, parents, governments, industry leaders, and experts in the fields of privacy, child safety, wellbeing and education to design better, safer products for kids and teens." 

For YouTube (via their blog blog.youtube/news-and-events/

  • YouTube default privacy settings for users aged 13 to 17 will be the “most private option available” (that only lets content be seen by the user and whomever they choose - teen users can make their content public by changing the default upload visibility setting)
  • YouTube will also start to remove “overly commercial content” from YouTube Kids" (for example, videos that focus on product packaging or encourages children to spend money)
  • YouTube will have “take a break and bedtime reminders" by default for all users 13-17. (Some adults could use that feature!)
  • YouTube will turn off autoplay by default for this age group

There are also changes for other parts of the platform, including search. 

Google will be introducing new policies that allow people under 18, or their parent or guardian, to request the removal of their image from Google Image results. Removing an image from Search doesn’t remove it from the web. They will also be turning on its SafeSearch, which aims to filter out explicit results, for all existing users under 18 and make it the default setting for teens who set up new accounts. The SafeSearch update will be rolled out “in the coming months,” according to Google. 

In other apps, Google will disable location history for all users under 18 without the ability to turn it on. A safety section in Play will show parents which apps follow Google's Families policies and disclose how they use the data they collect in greater detail.

Of course, ads are where Google makes its money, but it will "block ad targeting based on the age, gender, or interests of people under 18." 

 

It Is Way Past the Time to Update the Communications Act of 1996

social media
Image by Pete Linforth from Pixabay

If you have been using the Internet for the past 25 years, you know how radically it has changed. And yet, no comprehensive regulations have been updated since then.

The news is full of complaints about tech companies getting too big and too powerful. Social media is often the focus of complaints. We often hear that these companies are resistant to changes and regulations, but that is not entirely true. 

On Facebook's site concerning regulations, they say "To keep moving forward, tech companies need standards that hold us all accountable. We support updated regulations on key issues."

Facebook may be at the center of fears and complaints, but they keep growing. Two billion users and growing.

There are four issues that address that they feel need new regulations.

Combating foreign election interference
We support regulations that will set standards around ads transparency and broader rules to help deter foreign actors, including existing US proposals like the Honest Ads Act and Deter Act.

Protecting people’s privacy and data
We support updated privacy regulations that will set more consistent data protection standards that work for everyone.

Enabling safe and easy data portability between platforms
We support regulation that guarantees the principle of data portability. If you share data with one service, you should be able to move it to another. This gives people choice and enables developers to innovate.

Supporting thoughtful changes to Section 230
We support thoughtful updates to internet laws, including Section 230, to make content moderation systems more transparent and to ensure that tech companies are held accountable for combatting child exploitation, opioid abuse, and other types of illegal activity.

The Telecommunications Act of 1996 was the first major overhaul of telecommunications law in almost 62 years. Its main goal was stated as allowing "anyone [to] enter any communications business -- to let any communications business compete in any market against any other." The FCC said that they believed the Act had "the potential to change the way we work, live and learn." They were certainly correct in that. But they continued that they expected that it would affect "telephone service -- local and long distance, cable programming and other video services, broadcast services and services provided to schools."

And it did affect those things. But communications went much further and much faster than the government and now they need to play some serious catchup. It is much harder to catch up than it is to keep up.