Connect with us


New Freshness Algorithm for Featured Snippets

Darren Rogers



Google has just put in production a new algorithm displaying more recent content in its “Featured Snippets” when it considers that the search requires it (request called “hot”) …
Google announced today the implementation of a new algorithm for its Featured Snippets for queries to search for “hot” content.

Google takes the example of a search on “tax brackets”, explaining that this type of request is rather intended to get the latest figures published and not those of previous years.

Of course, all searches do not necessarily require a “fresh” answer. All requests are not “hot” (news). The new algorithm therefore tries to detect “hot” requests by providing a Featured Snippet that is as up to date as possible, without affecting this criterion for “cold” requests that do not require freshness.

Nothing really new, because this concern has always been taken into account by the engine. This algorithm seems to be an evolution of the QDF system (Query Desserves Freshness) set up in 2007 and the Freshness Update of 2011, but more specifically adapted to the featured snippets.

The article does not say however whether this novelty has been implemented worldwide today or only in the United States at first. Anyway, it should quickly happen in our country if it is not already the case.

Continue Reading


Apple, Google pause reviewing audio recordings from voice assistants

Mandy Boyd



Following a complaint from a German organization, Apple, Google and Amazon all announced – to varying degrees – a suspension of their system of listening to conversations of users of their voice assistants.

A few weeks ago, Amazon and Google had admitted that they used subcontractors or employees to listen to certain conversations (less than 1% in general) of users of their voice assistants, in order to “improve their voice services.” Apple, it seems, did the same with Siri. The news shocked many observers, surprised that these companies can listen to potentially personal information without clear warning, or consent.

In this context, the Data Protection Authority of Hamburg has just announced that it has obtained a suspension of these practices that will apply for three months in the European Union for the voice assistant of Google, and in a definitive way for Apple, all over the world.

“These companies must transparently inform the persons concerned about the processing of voice commands, but also about the frequency and risks of nuisance tripping that they generate,” said Johannes Caspar, Commissioner for Data Protection and Privacy. freedom of information within the German authority.

Apple has indicated that it plans to put in place a prior authorization request system to its users to implement this listening process. Google has suspended its practices for 3 months from August 1 without explaining precisely what it intends to do to improve the situation. In the process, Amazon has announced that it will allow Alexa users to disable the listening system.

Everything is suspended for now. It remains to be seen, by the end of the year certainly, how these systems will be reintroduced in Europe to be more consistent with the RGPD.

Continue Reading


Twitter used personal data without express consent

Cam Potvin



Twitter announced Tuesday that since May 2018, personal data had been used for advertising purposes without the explicit consent of its users.

The social network said the mistake was corrected Monday and that an investigation is underway to determine how many people could be affected, advising the passage to users who use its service to review their settings for data sharing.

According to Twitter, the use of personal data was carried out in two ways. In the first case, information was transmitted to Twitter’s advertising partners when the user clicked on an advertisement for a mobile application.

This data leak started in May 2018, coinciding with the entry into force of the European Union’s General Data Protection Regulation (GDPR). The RGPD requires companies to disclose such leaks of personal data within 48 hours of discovery and provides significant fines for business violating the rules.

Monitoring on the web

In the second case reported by Twitter, the social network says it has presented targeted advertising to users by monitoring their browsing the web, even when it had been banned this collection of data.

Since September 2018, we have been able to show you advertisements based on what we have deduced from the use you make of your device, even if you have not given us permission. The data concerned remained in the hands of Twitter and did not concern information such as passwords or email addresses , explained the network in an explanatory note published on its online help center.

Once again, Twitter could be sanctioned under the GDPR in Europe, since the regulation requires web platforms to ensure the explicit consent of users to collect their data, including for advertising targeting purposes and when such collection is for third-party companies.

Twitter, which apologizes to those concerned, ensures that the necessary steps are taken to prevent this from happening again , and invites people to contact their data protection office via a form.

Continue Reading


Study: 48% of consumers use voice assistants for general web search

Mandy Boyd



Adobe has just published a new study on the use of voice enabled search in the United States.  The figures confirm that the advance of these technologies is very real

Every week or we come across another study on the phenomenon of voice assisted search, trying to convince us every day of the imminence of this new paradigm of communication with digital interfaces.

Today, it is now Adobe’s turn to offer a study of 1,000 users of these technologies in the United States. Here are the main results:

  • 94% of users believe that voice technology is easy to use and claim that it saves them more than time – it improves their quality of life.
  • But more than half of them say that the process of using voice technology is not intuitive: 49% say they do not always know where to start to do a voice task.
  • And while users are satisfied with the ability of voice assistants to work on multiple devices and respond to commands, less than half of them use voice technology on a daily basis.
  • People are already using voice for a wide range of simple tasks, including driving directions, phone calls, SMS, weather checking and music listening. But most say they would not use it for more complex tasks such as personal banking (61%) or travel booking (52%).
  • Many brands strive to design voice assistants who, like people, can hold conversations. But the respondents do not agree. Consumers are divided on whether voice technology should (51%) or not (49%) have human attributes, such as sympathy and humor, as it continues to evolve.
  • On average, users say that voice technology understands them and is accurate when given an order or asked a question only 69% of the time.
  • Nearly half of respondents (47%) said that they sometimes feel uncomfortable talking to a machine.
  • Smartphones are already the most used device for voice technology (85%) – far ahead of smart speakers (39%) – indicating that voice and screens are a winning duo. And most users (80%) agree that visuals would allow them to use voice technology for a wider variety of tasks, with 83% saying that a screen reiterating the command given to confirm understanding would be particularly useful. In short, indispensable screen!

According to this study, the main current uses of voice (mainly on smartphones) are:

  • Get Driving Directions: 52%.
  • Make a phone call: 51%.
  • Sending a text message: 50%.
  • Obtaining the weather: 49%.
  • Listening to music: 49%.
  • General search on the Web: 48%.
  • Setting various alarms: 41%.
  • Listen to the news: 27%.
  • Sending email: 17%.
  • Shopping: 16%.

On the other hand, it was also asked what potential uses voice enabled search could offer in the coming years. Here are the answers:


Of course, this study is specific because it concerns only respondents located geographically in the United States and the target is characterized by people already using voice search. But this is an additional study to add to the list already provided of the various polls already available.

Continue Reading