No one can leave social media (WeChat, Weibo, Facebook, etc.). The circle of friends is one of the most important channels to maintain the outside world, and most people are heavy users of social media. However, have you ever thought about our unconscious behaviors on social media (such as likes, collections, uploading avatars, status updates, reposting, etc. In fact, it revealed some secrets and privacy at a deep level.
Combing some achievements of social media data mining, you will find that the combination of big data, behavioral traces and social media shows unlimited opportunities and possibilities. For example, Facebook-based mining can reveal your IQ, neuroticism, life satisfaction, religious belief, single or not, obesity, postpartum depression, topic preference and public opinion monitoring.
First, don't praise easily-it will reveal your personality secret.
I read an article recently, which is very interesting.
A recent study in the Proceedings of the National Academy of Sciences (PNAS) shows that you should not praise it easily on social networking sites, because praise can reveal some of your more personal personality characteristics. Based on the likes on social networking sites (such as Facebook), you can calculate your personality characteristics (big five personality), and the results calculated by the algorithm are more accurate than those judged by your friends, classmates and relatives.
Research process. Specifically, the author invited 86,000 volunteers from Facebook to take a personality test and collect their "likes" data (that is, what posts or content they like). The relatives and friends of the subjects were invited to participate in the test to evaluate the personality of the subjects. In this way, there are three personality data about the subjects, one is self-evaluation, the other is evaluation by relatives and friends, and the other is the result based on favorable data. The research results show that the personality tendency index obtained by the algorithm is more accurate than the good judgment of relatives and friends.
Like and special. When we like it, we want to show our Facebook friends our positive attitude towards certain content (including status updates, photos, books, products and music). However, at the same time, "liking" behavior also exposes a lot of your private information, sensitive characteristics, personality preferences and behavioral tendencies. Such as religious belief, political views, sexual orientation and drinking. Moreover, usually a praise (such as liking a share) can produce more accurate predictions. Some conclusions are as follows.
Specifically, with only about 10 likes, the computer can judge your personality more accurately than colleagues. Through 70 likes, the computer can surpass its friends. 140 "likes" more than their families (parents, brothers and sisters). 300 likes can "beat" your other half.
Second, some other interesting cases and applications related to social media.
(1) Are you still single?
Status update and single judgment. When I found Sister Gao Yan, the first thing my friends thought was whether she was single or not. Tell you a gospel, Facebook has helped you solve this problem.
Through millions of updated data, Facebook analyzes the text based on the updated content and identifies some characteristic variables, thus judging and predicting when users will change from single state to love state. In addition, the Facebook joint venture company will apply the research conclusions to business, that is, put relevant product advertisements on the homepage of personal web pages. For example, provide products for users in love (buy one get one free, couple discount, etc. ), or provide products to lovelorn users (paper towels, Ben&; Jerry ice cream, etc.
(2) Content collection and personality judgment
In the circle of friends, I often collect some content and documents. In fact, you can also judge your highly sensitive personal characteristics through your collection. Such as sexual orientation, life satisfaction, intelligence, emotional stability, religious belief, excessive drinking habits, age, gender, race, political views, etc.
(3) Friend recommendation in social media
Everyone on Facebook needs to upload their own avatar. What can Facebook do based on the user's avatar? Face recognition technology.
Face recognition refers to analyzing the relative position, size and shape of five senses according to facial features, retaining the most prominent features, and transforming image data into matrix data for face recognition.
After face recognition of the user's head, Facebook will search the whole internet to see where these photos have appeared, see if they have appeared in other people's photos, guess who you might know and recommend some friends with two degrees and three degrees.
On the other hand, Facebook will also match photos with company websites and online dating websites, or look for more information about you from newly published blogs or articles. So as to understand your occupation, interests, preferences, and behavioral tendencies.
(4) Who is this stranger?
Google glass and social media. I saw a video before, which described a little friend who met a stranger after wearing Google Glass, and quickly found a lot of information about this stranger through Google Glass. Think about it, it can be achieved.
Google glass needs to cooperate with Facebook. Google has developed a set of name tag technology. Based on this technology, when meeting a stranger, Google Glass will take a photo, transmit the person's face and throw it on Facebook for facial recognition. So as to match the occupation, age, birthday, dynamics, circle of friends, diary, status and recent interests of strangers.
(5) Who is the most suitable candidate?
Now some companies use social media content to screen potential employees. Recruiters capture the personal information of Facebook through personal data, such as Facebook, and judge and predict the intelligence level and emotional stability of individuals according to these information and the characteristics of individuals in the network environment. As a reference standard for employee recruitment.
(6) Long time no see, have you lost weight?
Analysis of avatar change. On Facebook, everyone's head is different. Every once in a while, everyone will adopt a new avatar or take a selfie. When you update your avatar, Facebook is thinking about another thing, that is, they use face recognition technology to judge whether there is a difference between the front and back avatars, that is, whether you have gained weight.
If the result of image analysis shows that you are really fat and your gender happens to be female, Facebook is likely to join forces with some weight loss, fitness and health care companies to accurately push some discount coupons to your personal page.
(7) Postpartum depression
More and more companies care about the welfare and living conditions of employees, and intervene in time to make you happy in your work and life.
A company judges whether female employees are at risk of falling into postpartum depression by analyzing their posts and topics on social media (such as Weibo, Facebook and Twitter). By capturing the topics and speeches of female employees on social media in the weeks before and after production, the verbs they use are extracted.
Analysis shows that mothers who are prone to depression often use words that imply "anxiety" and "unhappiness". A speech usually contains some derogatory terms, such as "disappointment", "pain" and "disgust", which implies the madness of disappointment.
(viii) Social media and disaster relief road map
When Typhoon 20 13 swept through the Philippines, people drew a road map for disaster relief through PO's photos on social media.
When disasters happen, people often post photos, texts and videos of their experiences to social media. Researchers analyze and mine photos uploaded by people. For a photo of the disaster, "What is the loss? Click the button to confirm, None, Medium and Severe ". For Facebook or Twitter text messages, judge the text types, such as "seeking help", "infrastructure damage", "population loss" and "related but other".
Then, by confirming where the data comes from in the Philippines, it can be located by GPS sensing information carried by photos or words, thus drawing a disaster map.
Through this disaster relief map, we can know who needs what, who needs food, where the bodies are, where there are displaced people, where the disaster is more serious and what hospitals are least damaged. So as to provide first-hand and freshest information for disaster relief.
(9) Public opinion and topic monitoring
At present, many companies are tracking and monitoring brand related words (products, brands, competitors, brand spokespersons, policy trends, etc.). ), and analyze emotions and topics to understand how people discuss companies and their products on the Internet.
Social media is also an early warning system. When consumers are dissatisfied with products or services, they will vomit on social media, and enterprises will quickly capture this information and improve products and services to control the spread and spread of negative news.
Third, digital trace and intelligent life.
"Big data" is everywhere, and "intelligence" is sexy. When "big data" and "intelligence" blend and ferment, everything becomes infinite.
What do you want to express through these cases of social media? About what time?
First, whether you are conscious or not, many of your actions will leave digital traces. Text (Weibo), video (Youku Tudou, etc. ), audio (customer service phone), pictures (friends circle), sensor data (bracelet, mobile phone built-in sensor), etc.
Second, don't deny that machines and algorithms sometimes know you better.
Third, the algorithm is rational, intelligent and sexy. Moreover, algorithms and intelligence go hand in hand.
Fourthly, by extension, based on data mining, computers can infer the psychological characteristics of you and me. Furthermore, if the computer can make a matching response according to the individual's personality characteristics, it will truly become an intelligent product with emotional and social skills.
I don't know if you have seen the movie "She and her predecessor Machina". Dream Girl based on artificial intelligence in her body is definitely your perfect companion.
Fifth, last but not least, at least I have worked hard to sort out so much information, and remember to praise it after reading it.