Connect with us


Machine learning to re-identify anonymized data



Researchers from Imperial College London and the Catholic University of Louvain (UCLouvain) in Belgium explain in an article that it is possible to re-identify individuals by reverse-engineering a sample of anonymous data to the using machine learning techniques.

The main solution for sharing data while preserving the privacy of individuals is to remove from the dataset the registration of direct identifiers such as name and email address, and to share only a portion of these. data. For example, consider the case of several women in their thirties living in Brisbane, Australia. Looking for these demographics in a sample of anonymized data should, therefore, theoretically find any number of people. “But the problem is that it does not work,” the researchers said. With only a few more attributes, a record is quickly becoming more exceptional.

The statistical model of researchers quantifies the probability that an attempt at re-identification will succeed, even with a “very incomplete” dataset. For example, according to an online tool that supports their demonstration, with only gender, marital status, date of birth, and postal code, there is a 86% chance of correctly identifying an individual in any set of anonymized data. “This is information quite commonly requested by companies,” said Dr. Yves-Alexandre de Montjoye, lead author of the article “Estimating the success of re-identifications in incomplete datasets using generative models” published in Nature Communications. According to this publication, 99.98% of Americans were correctly re-identified in any anonymized data set available using only 15 characteristics, including age, gender, and marital status.

Any facts, grouped together

Validated on 210 datasets from demographics and surveys, researchers say their technique – which uses Gaussian copulas to model uniqueness – shows that “even very small sample sizes are not enough to prevent re-identification and protect the data “. In fact, “contrary to popular belief, the sampling of a dataset does not offer plausible deniability and does not effectively protect the privacy of individuals,” added Yves-Alexandre de Montjoye.

Shared and exchanged datasets often contain many attributes. For example, the Experian data broker sold Alteryx access to an anonymized dataset containing 248 attributes per household for 120 million Americans. “Even though there are a lot of people in their thirties, male, living in New York, there are fewer people who were born on January 5, drive a red sports car and live with two children, two girls, and a dog. There is probably only one individual meeting these criteria, “said Dr. Luc Rocher, co-author of the article. There are few protections against such attempts to re-identify – although the Australian federal government has already considered criminalizing the re-identification of Commonwealth datasets disclosed as part of its open data program.

A risk of re-identification downplayed by governments

The sale of a sample of anonymous data means that they are no longer subject to data protection regulations – such as the EU General Data Protection Regulation (GDPR) and the Consumer Privacy Act (CCPA). of the State of California – and therefore they can be freely used and sold to third parties such as advertising agencies and data brokers. In its “Disidentification” Guide published last year, the Office of the Privacy Commissioner of Australia states that sampling creates “the uncertainty that a particular person is included in the data set”. But, according to Mr. de Montjoye, “businesses and governments minimize the risk of re-identification by arguing that the datasets they sell are still incomplete. Our findings show that this might not be helpful. “

There are many examples of supposedly anonymous data sets that were later disclosed and reidentified. In 2016, journalists re-identified public figures in a set of anonymized data corresponding to the browsing history of 3 million German citizens they acquired for free from a data broker. They were able to rediscover a judge’s pornography preferences and the drugs used by a member of Parliament. In the same year, researchers at the University of Melbourne were able to decrypt identification numbers used by service providers in a 10% sample of medical billing records published by the Australian Department of Health. This sample could allow Medicare service providers to re-identify data in the dataset.

Stricter rules for anonymous sharing of requested data

A year later, the same researchers showed how patients could also be re-identified by linking unencrypted portions of their records to known information about the person. “A few banal facts taken together are often enough to isolate an individual,” remark Culnane et al. Imperial College and UCLouvain researchers have called for stricter rules on anonymous sharing of data. “The purpose of anonymisation is to facilitate the use of data for the benefit of society. This is extremely important, but it should not and should not be at the expense of the privacy of individuals, “said Professor Julien Hendrickx, another co-author of the article. “It is essential to apply robust anonymization standards and to take into account new threats such as those shown in this document,” he added.

Continue Reading


  1. Save The OA

    August 11, 2019 at 8:51 pm

    Thats what an angel is. Dust pressed into a diamond by the weight of this world. #TheOA needs your help. #SaveTheOA

  2. Save The OA

    August 13, 2019 at 1:29 pm

    Old Knight and psychic octopi are real! #TheOA needs your help. #SaveTheOA

Leave a Reply

Your email address will not be published. Required fields are marked *


Amazon’s new ,Fire HD 10 Tablet , With low cost !!!




The new tablet will come in a regular edition and a kids edition, but the core technology remains the same on both. Let’s get the specs out of the way: it has a 10.1-inch FHD touchscreen, an updated octa-core processor, 2GB of RAM, and either 32GB or 64GB of storage that you can expand with its microSD card.

Amazon increased the battery life to 12 hours, up from the 10 hours of life the previous Fire HD10 could get. The tablet can charge fully in four hours thanks to its new USB-C port. This replaces the microUSB port that the previous model had and is arguably the most exciting thing about the Fire HD 10. It’s the first Amazon device with a USB-C port for charging, giving us hope that we could see the company ditch microUSB altogether in the future. The price is $ 149 .

On the software side, the Fire HD 10 runs Android P and has all of the features the previous model did, including hands-free Alexa (should you choose to enable that feature). Amazon also added a new picture-in-picture feature that minimizes certain video apps, so you can still watch video content while using other programs like your calendar, a notes app, and others. The feature will be limited by the supporting video apps, though, and Amazon Prime Video and Netflix currently work with it.

The Fire HD 10 Kids Edition has the same UI as Amazon’s existing tablets for kids, making it more accessible and easier to use for children. The tablet also comes with a kid-friendly case, an extended warranty, and a one-year complimentary FreeTime Unlimited.

Amazon has added a lot to FreeTime Unlimited over the years and continues to do so. The kid-focused, $4.99-per-month ($2.99-per-month for Prime members) content subscription includes thousands of shows, movies, books, audiobooks, and other media that kids can consume without parents needing to filter out inappropriate materials.

In addition to being available on the kids’ versions of Fire tablets, Amazon is also bringing FreeTime Unlimited to Fire TV Sticks starting today. That means kids can watch shows and movies through FreeTime Unlimited, as well as Prime Video content and content that parents deem suitable from their personal libraries. All of this will be in one place on TVs with connected Fire TV Sticks.

Continue Reading


Robot Chef That Can Prepare Your Dinner




It can do a very pleasant crab bisque in less than 30 minutes

Ever since Americans were introduced to Rosie, the beloved robot maid on The Jetsons, way back in the 1960s, robotic household help has been the ultimate in futuristic dream products.

A new product from Moley Robotics might bring that future one step closer, as the company unveiled a robot chef on Tuesday at Hannover Messe, a trade fair for industrial technology in Germany. Comprised of two robotic arms in a specially designed kitchen, which includes a stove top, utensils and a sink, the device is able to reproduce the movements of a human chef in order to create a meal from scratch. The robot learns the movements after they are performed by a human chef, captured on a 3D camera and uploaded into the computer.

A few weeks before the robot chef was unveiled, Moley invited TIME to check out the robot and test its fare. In less than half an hour, the robot made a crab bisque, based on the recipe and technique of Tim Anderson, winner of 2011’s season of MasterChef in the U.K., who is working with Moley to develop the kitchen. From selecting the right heat level on the stove-top to adding the pre-arranged ingredients at just the right moment to operating a small mixer, the robot arms made the soup from scratch. It even plated up the soup, including scraping the bottom of the ladle against the rim of the saucepan in order to prevent drips.

Why crab bisque? “Crab bisque is a challenging dish for a human chef to make, never mind a robot,” explains Anderson. “If it can make bisque, it can make a whole lot of other things.” When asked if he feels at threatened by seeing a machine expertly recreate one of his recipes, Anderson is somewhat surprisingly on the side of the technology. “Some people ask if this is going to put my out of a job. This has already given me a job.”

Comparing the robot to cookbooks and YouTube tutorials by professional chefs, Anderson says, “I think it’s going to help people build brands.” The aim is to have professional chefs record themselves cooking their own recipes so that the robot will be able to mimic the techniques and replicate the dish. Anderson envisions people learning how to make a variety of dishes by watching their robots in action. “It’s changed the way I think about cooking,” he says.

Moley, which was founded by computer scientist Mark Oleynik, has partnered with the London-based Shadow Robot Company, which developed the kitchen’s hands. Twenty motors, two dozen joints and 129 sensors are used in order to mimic the movements of human hands. The robotic arms and hands are capable of grasping utensils, pots, dishes and various bottles of ingredients. Olyenik says that the robot hands are also capable of powering through cooking tasks quickly, though they’ve been designed to move quite slowly, so as not to alarm anyone watching it work.

Sadly for vegetarians, like Shadow Robot’s managing director Rich Walker, crab bisque is the only dish the robot is currently able to make. However, the company plans to build a digital library of 2,000 recipes before the kitchen is available to the wider public. Moley ambitiously aims to scale the robot chef for mass production and begin selling them as early as 2017. The robotic chef, complete with a purpose-built kitchen, including an oven, hob, dishwasher and sink, will cost £10,000 (around $15,000). Yet that price point will depend on a relatively high demand for the kitchen and it’s still unclear how large the market is for such a product at the moment.

Dan Kara, a robotics analyst for U.S.-based ABI Research, a market intelligence company that specializes in emerging technology, tells TIME that the household robots that have found a market tend to be smaller devices that tackle tedious chores. “Successful products for the home that I’ve seen have been floor-cleaning — sweepers and vacuums — and pool cleaners and lawnmowers,” he says, noting that people tend to favor robots that tackle tasks they don’t want to do “because it’s boring or repetitive.” Another key factor of a product’s success is affordability. “Once get above a certain price, the number of people using them drops right off.”

A robotic chef, however, “just seems like a bridge too far at this time,” though Kara pointa out that he isn’t familiar with Moley’s kitchen or its specific technology.

Which isn’t to say that a robot chef wouldn’t have interested buyers. The robotics industry is growing and the Boston Consulting Group has estimated that spending on robots could “jump from just over $15 billion in 2010 to about $67 billion by 2025.”

But there is still work to be done on Moley’s kitchen before it would be an even remotely practical, albeit pricey, purchase. As the robot doesn’t have any way to see, it’s unable to locate an ingredient or utensil that might be moved or knocked out of place. It also can’t chop or prep food yet, so it must use prepared ingredients that are meticulously laid out. The company is working on improving the robot’s functions and expanding its capabilities, but as Oleynick admits, “it will have some limits because nothing can replace human touch.”Source:

Continue Reading


Social Media Marketing




“Social media has become a primary tool for higher levels of fan engagement, directly driving lead generation through interaction and content sharing that is especially relevant to media companies. Social Media Marketing: The Next Generation of Business Engagement deconstructs the tools and techniques, showing you how to apply social technology to your business.”

The Next Generation of Business Engagement

Social technologies, on a mass scale, connect people in ways that facilitate sharing information, thereby reducing the opportunities for marketplace exploitation—whether by charging more than a competing supplier for otherwise identical goods and services or charging anything at all for products that simply don’t work. Sunlight is a powerful disinfectant, and the collective knowledge that powers the Social Web is the sunlight that shines in these new connected marketplaces. The Social Web dramatically levels the playing field by making information plentiful, just as it also levels businesses and organizations that operate on the principles of making information scarce. The Social Web exposes the good, the bad, and the ugly, simultaneously raising up what works and putting down what doesn’t without regard for the interests of any specific party. Web 2.0 technologies—expressed through social CRM, vendor relationship management, collective ideation, customer-driven support forums, and communities where participants engage in all forms of social discourse—act together to equalize the market positions of suppliers, manufacturers, business and organizational leaders, customers and stakeholders. To again quote Sir Tim Berners-Lee, “If misunderstandings are the cause of many of the world’s woes, then (we can) work them out in cyberspace. And, having worked them out, we leave for those who follow a trail of our reasoning and assumptions for them to adopt, or correct.”

So whether supporting Unilever, P&G, and Nestlé, all working with Greenpeace to ensure supplier compliance in the use of sustainable palm oil and thereby reducing environmental damage in no-longer “far away” places like Malaysia, or just making someone’s day run a little more smoothly by preventing a coffee stain through a simple innovation like Starbucks’ “no splash” stirring stick, the businesses and organizations embracing social technologies are delivering better solutions—developed through direct collaboration with customers and stakeholders—to the world’s woes however large or small they may be. Contemporary businesses, cause-based organizations, and governing authorities are increasingly meeting the challenge of “opening up” and operating with their customers and stakeholders—often through a similarly empowered and connected workforce—to deliver self-evident value that gets talked about. For these entities, their customers, suppliers, and stakeholders are the new source of future innovations and “marketing,” and therefore also the drivers of long-term growth and success. This is what social business is all about.

Social Business (Feedback)

For a lot of organizations—including business, nonprofits, and governmental agencies—use of social media very often begins in Marketing, public communications, or a similar office or department with a direct connection to customers and stakeholders. This makes sense given that a typical driver for getting involved with social media is a slew of negative comments, a need for “virality,” or a boost to overall awareness in the marketplace and especially in the minds and hearts of those customers increasingly out of reach of interruptive (aka “traditional”) media. In a word, many organizations are looking for “engagement,” and they see social media as the way to get it. The advent of Web 2.0 and the Social Web is clearly a game-changer, on numerous fronts. Given the rush to implement, and the opening focus on marketing specifically versus the business more holistically, many “social media projects” end up being treated more like traditional marketing campaigns than the truly revolutionary ways in which a savvy business can now connect with and prosper through collaborative association with its customers. As a result, the very objective—engagement, redefined in a larger social context—is missed as too many “social media campaigns” run their course and then fizzle out.

everyone. The collaborative technologies that now define contemporary marketplaces—technologies commonly called “social media,” the “Social Web,” or “Web 2.0”—offer a viable approach to driving changes in deeper business processes across a wide range of applications. There is something here for most organizations, something that extends very much beyond marketing and communications. This chapter, beginning with the Social Feedback Cycle, provides the link between the basics of social media marketing and the larger idea of social technologies applied at a “whole-business” level. As a sort of simple, early definition, you can think of this deeper, customer-driven connection between operations and marketing as “social business.”

Beginning with the emergence of Web 2.0 technologies—the set of tools that make it easy for people to create and publish content, to share ideas, to vote on them, and to recommend things to others—the well-established norms of business marketing have been undergoing a forced change. No longer satisfied with advertising and promotional information as a sole source for learning about new products and services, consumers have taken to the Social Web in an effort to share among themselves their own direct experiences with brands, products, and services to provide a more “real” view of their research experience. At the same time, consumers are leveraging the experiences of others, before they actually make a purchase themselves. The impact on marketing has been significant, to say the least.

Continue Reading