April 6, 2017 at 23:53Look at yourself through the eyes of Zuckerberg and help his competitors
Sometimes you write, you write in social networks on an important topic, and the reaction is zero. Doubt involuntarily creeps in, or am I writing. And so do I write as it should. A look from the outside on your behavior in social networks can reveal those of your qualities and features of behavior that you know, but consider secondary. In this sense, Zuckerberg has more truthful assessments of your personality. Not always to the smallest detail, but always about the essence of your character and behavior. If only he has the right algorithms for converting your actions into traits of character traits. Psychologists have such algorithms a long time ago. We are saved only by the fact that on the other side of the monitor we are more willing to sell something than to manage us.
However, knowing how you behave in the eyes of algorithms is a good thing. If only because that is how you and one of your significant friends on social networks see you, and you don’t even suspect about it. The ability to see your digital selfie is given by the dataselfie.it program and its extensions for the chrome browser.
Data Selfie collects information about what you click (likes and links), what you print, what and how long you view. Based on the monitoring results, the program provides the following information:
The program interface is so-so. Obviously not for commercial use. Developers do not plan to sell their service, are guided by good intentions and claim that they do not store monitoring data anywhere other than the user's computer.
The video shows what the interface and monitoring reports look like.
For those who have doubts about the integrity of the monitoring, the developers came up with an argument and send the curious to the source code of the program on github . Anonymous data is transmitted to the server for processing, of course. Including third parties whose servers analyze user behavior. And here is quite interesting. The main machine learning algorithm used by the service is Apply Magic Saucefrom developers at the University of Cambridge Psychometric Center. This is a network of strategic research in the field of psychological, professional, clinical and educational assessment. Part of this network is Cambridge Analitica, known as the Brexit demiurge and the election of Donald Trump. In addition to belonging to a common research network, both Cambridge programs have a common methodology for psychological stratification. The monitoring uses the so-called “ocean method” (OCEAN - this is the first letter in English). The "Big Five" of evaluation parameters include: openness (how much are you ready for a new one), honesty (how much are you a perfectionist), extraversion (how do you feel about society), goodwill (how friendly and willing to cooperate) and neuroticism (how easy you are make one mad).
On Facebook, the Who Are You by Psychotype tests are very common. The most famous and relatively ancient - the MyPersonality application - was launched in 2008 by Pole Mikhail Kosinsky, a student at Cambridge University, who later became a leading employee of Cambridge Analytica, and now Stanford. Zuckerberg and Co. himself are, of course, out of competition and he has long ago laid out our fears, hopes and manner of communication on the right shelves. But the British (and Cambridge is England) do not intend to lag behind in the game “to sell correctly - to play on the nerves”.
In full accordance with machine learning methods, this technology allows us to evaluate the preferences of Facebook users by comparing the structure of their likes and reposts with other data depending on the purpose. The data of tens of millions of network accounts is used, which allow us to identify the most relevant combinations of user characteristics (gender, interests, religion, political views, personality type, orientation, marital status and others) in the so-called percentiles (with what probability the user belongs to one or another category )
It’s clear why you need this company to analyze your data through the Data Selfie website for free, albeit depersonalized. Her engineers thus refine the correlations and check the accuracy of the algorithms. To sell them as part of advertising or elective targeting is a technical matter.
Speech by the Director of Cambridge Analytica on the principles of microtargeting in election campaigns:
The first thing that leads to awareness of one’s openness is paranoia. You will become more aware of your actions, data, put anonymizer and vpn. You make life difficult for yourself, but not for Big Brother Network. You will hide 20% of the information from him and from scammers (which is more important than IMHO), but you can manage you by the remaining volume of your interests. Better use this information to correct impressions and cut off empty slides on the net.
Like, for example, I’m completely crazy to drive AI crazy when communicating on Facebook. Since there are quite a lot of friends, and Zuckerberg shows their posts using an unknown algorithm, I set my own rules for viewing posts. Firstly, I created about 20 lists (by projects, interests and activities) and distribute my friends there. In the lists, I see all the posts of all the participants in the list and there is no need to look for them in the feed. For two consecutive months, a participant cannot be on the same list. Secondly, for unallocated friends on the lists for a month I remove the subscription function when I see it in the tape, and at the beginning of each next I activate the universal subscription again. New friends are always in priority impressions at first, but also flow into lists and friends, but without showing in the stream. I don’t intend to exaggerate the complexity of my method,
However, knowing how you behave in the eyes of algorithms is a good thing. If only because that is how you and one of your significant friends on social networks see you, and you don’t even suspect about it. The ability to see your digital selfie is given by the dataselfie.it program and its extensions for the chrome browser.
Data Selfie collects information about what you click (likes and links), what you print, what and how long you view. Based on the monitoring results, the program provides the following information:
- Your position on political, religious, environmental, social and resonant issues;
- Your activity in chronological order indicating the motives and accents to which you have devoted the most attention;
- 10 of your best friends;
- 10 most significant resources for you;
- A list of keywords summarizing your interests;
- List of characters and organizations to which you pay the most attention;
- The relevance of your personal focus of interests as a reflection of the ease of diverting your attention to the current news agenda;
- Your psychological portrait in 5 directions: openness, honesty, sociability, suppleness and anxiety.
The program interface is so-so. Obviously not for commercial use. Developers do not plan to sell their service, are guided by good intentions and claim that they do not store monitoring data anywhere other than the user's computer.
The video shows what the interface and monitoring reports look like.
For those who have doubts about the integrity of the monitoring, the developers came up with an argument and send the curious to the source code of the program on github . Anonymous data is transmitted to the server for processing, of course. Including third parties whose servers analyze user behavior. And here is quite interesting. The main machine learning algorithm used by the service is Apply Magic Saucefrom developers at the University of Cambridge Psychometric Center. This is a network of strategic research in the field of psychological, professional, clinical and educational assessment. Part of this network is Cambridge Analitica, known as the Brexit demiurge and the election of Donald Trump. In addition to belonging to a common research network, both Cambridge programs have a common methodology for psychological stratification. The monitoring uses the so-called “ocean method” (OCEAN - this is the first letter in English). The "Big Five" of evaluation parameters include: openness (how much are you ready for a new one), honesty (how much are you a perfectionist), extraversion (how do you feel about society), goodwill (how friendly and willing to cooperate) and neuroticism (how easy you are make one mad).
On Facebook, the Who Are You by Psychotype tests are very common. The most famous and relatively ancient - the MyPersonality application - was launched in 2008 by Pole Mikhail Kosinsky, a student at Cambridge University, who later became a leading employee of Cambridge Analytica, and now Stanford. Zuckerberg and Co. himself are, of course, out of competition and he has long ago laid out our fears, hopes and manner of communication on the right shelves. But the British (and Cambridge is England) do not intend to lag behind in the game “to sell correctly - to play on the nerves”.
How does anonymous analytics work?
In full accordance with machine learning methods, this technology allows us to evaluate the preferences of Facebook users by comparing the structure of their likes and reposts with other data depending on the purpose. The data of tens of millions of network accounts is used, which allow us to identify the most relevant combinations of user characteristics (gender, interests, religion, political views, personality type, orientation, marital status and others) in the so-called percentiles (with what probability the user belongs to one or another category )
It’s clear why you need this company to analyze your data through the Data Selfie website for free, albeit depersonalized. Her engineers thus refine the correlations and check the accuracy of the algorithms. To sell them as part of advertising or elective targeting is a technical matter.
Speech by the Director of Cambridge Analytica on the principles of microtargeting in election campaigns:
The first thing that leads to awareness of one’s openness is paranoia. You will become more aware of your actions, data, put anonymizer and vpn. You make life difficult for yourself, but not for Big Brother Network. You will hide 20% of the information from him and from scammers (which is more important than IMHO), but you can manage you by the remaining volume of your interests. Better use this information to correct impressions and cut off empty slides on the net.
Like, for example, I’m completely crazy to drive AI crazy when communicating on Facebook. Since there are quite a lot of friends, and Zuckerberg shows their posts using an unknown algorithm, I set my own rules for viewing posts. Firstly, I created about 20 lists (by projects, interests and activities) and distribute my friends there. In the lists, I see all the posts of all the participants in the list and there is no need to look for them in the feed. For two consecutive months, a participant cannot be on the same list. Secondly, for unallocated friends on the lists for a month I remove the subscription function when I see it in the tape, and at the beginning of each next I activate the universal subscription again. New friends are always in priority impressions at first, but also flow into lists and friends, but without showing in the stream. I don’t intend to exaggerate the complexity of my method,