Government Data Mining - Spotlight #309

Good Monday morning. It’s September 16th. The Global Climate Strike begins Friday. We believe in science. Climate change is real, accelerating, and happening now according to the scientists—no matter what any politician says. Hundreds of climate actions will be led by student activists who are also striking from school. Find a location near you.

Today’s Spotlight takes about 5 minutes to read. Want to chat about something you see here? Press your email reply button or click the silver “Write George” button below.

2. News To Know Now

1. State and federal government regulators are talking to Google competitors in the search engine market as part of their official probe into the company. As the state attorneys general combine their efforts—50 strong and only missing California and Arkansas—the House investigation into Big Tech continues. You’re not imagining things if this feels similar to government technology probes into AT&T in the 1970s and Microsoft in the 1990s. Here’s a Reuters explainer article about the Google probes.

2. Amazon, a long-time target of diverse politicians such as President Trump and Senator Bernie Sanders (I-VT), is also under scrutiny by the Federal Trade Commission. Bloomberg is reporting that a team of attorneys and economists are interviewing third-party merchants about their relationship with Amazon versus other platforms like eBay and Walmart.

3. PayPal has suspended a KKK donations account although it won’t address specifics, the BBC reports. PayPal previously committed to reviewing its relationship with any account managed by an organization promoting racist views.

3. Government Data Mining Update

What we once called Big Data is now simply data. The unrelenting stream of information is captured by systems of every size. A 2017 IBM study posited that ninety percent of the information on the Internet had been created in the previous two years.

And that’s online. Imagine the databases created by organizations at work, recreation, and municipal governments. Our government data mining analysis covers four areas over four weeks.

1. Facial recognition’s growth – this week
2. Ancillary data from DNA testing and app use.
3. National and local algorithms to make sense of all the data.
4. Extensions into areas like personal health records and trackers.

Observers could spend every working minute analyzing facial recognition to stay updated with its constant changes. For example, Amazon recently announced a change to its Rekognition software that “improved accuracy for emotion detection (for all 7 emotions: ‘Happy’, ‘Sad’, ‘Angry’, ‘Surprised’, ‘Disgusted’, ‘Calm’ and ‘Confused’) and added a new emotion: ‘Fear’. Lastly, we have improved age range estimation accuracy; you also get narrower age ranges across most age groups.”

Somehow Amazon is still working on age estimation accuracy, but can detect fear.

Facebook also announced new privacy settings for DeepFace, its facial recognition software. That sounds nice, but remember that DeepFace is believed to be the largest facial recognition database in the world thanks to the 250 billion photos that have been voluntarily uploaded to Facebook. The company claims that it beats the FBI’s facial recognition programs with 15% more accuracy.

Google’s Face Match algorithm now makes use of a camera in its Nest Hub smart home display, which is a nice way of saying that Google’s thermostat and light controlling gizmos point an always-on camera at your living space. You can learn more about that in CNet’s excellent “Google collects face data now. Here’s what it means and how to opt out.“

The race to get this facial data isn’t only to sell you more stuff although that’s certainly helpful. Live Nation and its Ticketmaster subsidiary has said that it will use facial recognition at live events. Not so fast, say some artists like the aptly named Rage Against The Machine.

More than half of U.S. adults trust law enforcement agencies to responsibly use facial recognition, according to Pew Research. The approval rating drops to 36% for technology companies and 18% for advertisers. California lawmakers sent a bill last week to Governor Gavin Newsom that would ban state and local police from using facial recognition software on their body cameras.

Next week is Part 2 of Government Data Mining: Ancillary data from apps

Government Data Mining Update

And that’s online. Imagine the databases created by organizations at work, recreation, and municipal governments. Our government data mining analysis covers four areas:

1. Facial recognition’s growth – this week
2. Ancillary data from DNA testing and app use
3. National and local algorithms to make sense of all the data
4. Extensions into areas like personal health records and trackers.

Observers could spend every working minute analyzing facial recognition to stay updated with its constant changes. Amazon recently announced a change to its Rekognition software that had “improved accuracy for emotion detection (for all 7 emotions: ‘Happy’, ‘Sad’, ‘Angry’, ‘Surprised’, ‘Disgusted’, ‘Calm’ and ‘Confused’) and added a new emotion: ‘Fear’. Lastly, we have improved age range estimation accuracy; you also get narrower age ranges across most age groups.”

Somehow Amazon is still working on age estimation accuracy, but can detect fear.

Facebook also announced new privacy settings for DeepFace, its facial recognition software. That sounds nice, but remember that their DeepFace database is believed to be the largest in the world thanks to the 250 billion photos that have been voluntarily uploaded to Facebook. The company claims that it beats the FBI’s facial recognition programs with 15% more accuracy.

Google’s Face Match algorithm makes use of a camera in its Nest Hub smart home display, which is a nice way of saying that Google’s thermostat and light controlling gizmos point an always-on camera at your living space. You can learn more about that in CNet’s excellent “Google collects face data now. Here’s what it means and how to opt out.“

3. Government Data Mining Update

And that’s online. Imagine the databases created by organizations at work, recreation, and municipal governments. Our government data mining analysis covers four areas over four weeks.

Somehow Amazon is still working on age estimation accuracy, but can detect fear.

Next week is Part 2 of Government Data Mining: Ancillary data from apps

ment agencies to responsibly use facial recognition, according to Pew Research. The approval rating drops to 36% for technology companies and 18% for advertisers. California lawmakers sent a bill last week to Governor Gavin Newsom that would ban state and local police from using facial recognition software on their body cameras.

Next week is Part 2 of Government Data Mining: Ancillary data from apps

4. Two Big Search Studies Released

A new survey of nearly 1,600 search marketers confirms evolvingtrends for us to look at. SparkToro founder (and Moz co-founder) Rand Fishkin has done this analysis for years and is adept at creating the research documents and interpreting them.

Fishkin notes that there is consensus for the first time in fourteen years that quality content/relevance means more than either keywords or links. The debate between on-page keywords and off-page links still exists although having both is the ultimate goal.

Page speed and mobile friendliness are also ready for their closeups. At this point, any site not focused on both of these is missing the point of optimization. Even if more than half of your website’s visitors use a desktop computer to visit your site, the chances are excellent that Google’s mobile ranking algorithms control its visibility. And there is no way to opt in or opt out of mobile-first indexing.

Digital marketers Backlinko also released a study of five million search queries last week. The data confirms or quantifies a lot of existing knowledge. For example, title tags–the line that contains the link—have a 14% higher click-through rate when a question is used as the title. Pages with meta descriptions—the snippet underneath that link—get nearly more 6% clicks. And I’m here to tell you that meta descriptions that Sue or George write get a lot more than that.

Backlinko’s report says that the number one result for a search query (on that device in that location on that day) gets more than 30% of the clicks. It’s important to recognize that this is data from one study, not all visitors click any link, and we’re not sure that the methodology controlled for all variables like geography.

That said, there’s little doubt about the overall findings: there are more clicks on the upper links on page one than the lower links, there are more clicks on page one results than there are on page two, and there are way more on page two results than page three or higher.

Read SparkToro’s Ranking Factor Survey or Backlinko’s Click Through Rate Study.

5. Debugged: Ordering a Pizza from 911

You’ve probably seen an often-shared social media message that advises people to order a pizza if they call 911 and can’t talk in front of someone. This rumor has made the rounds for years and is not something police dispatchers are trained for. The LAPD says that “Operators are trained to recognize voice inflection, odd conversations that would indicate a dangerous situation, among other things.”

PolitiFact has the meme debugged here.

6. Also in the Spotlight

Amazon’s Alexa assistant will be the “beneficiary” of crowdsourced questions and answers from the public in a move that must terrify its brand safety monitors, according to Tech Times.

Alphabet, owner of a snappy video site called YouTube, has released data that says 55% of consumers use online video for shopping research, according to Search Engine Journal.

Alphabet’s Google unit, meanwhile, would like you to know that it’s ready for the fall viewing season with new carousels that describe movies and TV shows, including where to stream them. You can read their announcement here for a reminder that you can pay to have its snappy video site named YouTube to stream movies and TV shows for you.

7. Food for Thought: Employee Monitoring

Your organization has oodles of data that you can analyze beyond the typical Human Resources information like background checks, attendance and ratings. These range from employee emails, phone calls, and other communications, Internet browsing history, to facial recognition to social media information on quasi-professional sites like LinkedIn or industry-specific sites and apps.

How do you balance information overload that allows you to get a fair look at employees and what processes are in place to protect the organization when non-disclosed monitoring programs identify troubling information?

8. Protip: Google Photos Finds Text

One amazingly useful product of technology’s pursuit of facial recognition technology is the news that Google Photos can now search for text that appears in one of your images. Signs, printing on clothing, or documents you’ve taken a picture of maybe at a lecture or presentation can now be searched as text.

The new search combines with the photo’s facial recognition, timestamp, and geo data too.

9. Great Data: Baby’s First Year…of Sleep

Seung Lee collected data from the times his son slept during his first year. Then Lee knitted a blanket of the visualized data because of course he did.

See the one-of-a-kind baby blanket.

10. Coffee Break: Wisteria by The Met

During today’s coffee break, let The Met bring you on a photo tour of colorful wisteria trees.

Pretty and shiny trees here.