发布时间 : 星期一 文章2018年美国大学生数学建模竞赛F题优秀论文翻译及解读更新完毕开始阅读27dba17f11a6f524ccbff121dd36a32d7375c7c8
2018ICM
ProblemF: Cost of Privacy
Pervasivenessof, and reliance on, electronic communication and social media have become
widespread.One result is that some people seem willing to share private
information
(PI)about
their
personal
interactions,
relationships, purchases, beliefs, health,and movements, while others hold their privacy in these areas as very importantand valuable. There are also significant differences in privacy choices acrossvarious domains. For example, some people are quick to give away the protectionof their purchasing information for a quick price reduction, but at the sametime are unlikely to share information about their disease conditions or healthrisks.Similarly, some populations or subgroups may be less willing to give upparticular types of personal information if they perceive it posing a personalor community risk. The risk may involve loss of safety, money, valuable items,intellectual property (IP), or the person’s electronic identity. Other risksinclude professional embarrassment, loss of a position or job,social loss(friendships), social stigmatization, or marginalization. While a governmentemployee who has voiced political dissent against the government might bewilling to pay to keep their social media data
private, a young college studentmay feel no pressure to restrict their posting of political opinion or socialinformation. It seems that individual choices on PI protection and internet andsystem security in cyber space can create risks and rewards in elements offreedom, privacy, convenience, social standing, financial benefits, and medicaltreatment.
Isprivate information (PI) similar to private personal property (PP) andintellectual property (IP)? Once lawfully obtained, can PI be sold or given toothers who then have the right or ownership of the information? As detailedinformation and meta-data of human activity becomes more and more valuable tosociety, specifically in the areas of medical research, disease spread, disasterrelief, businesses (e.g. marketing, insurance, and income), records of personalbehaviors, statements of beliefs, and physical movement, these data anddetailed information may become a valuable and quantifiable commodity. Tradingin one’s own private data comes with a set of risks and benefits that maydiffer by the domain of information (e.g. purchasing, social media, medical)and by subgroup (e.g. citizenship, professional profile, age).
Canwe quantify the cost of privacy of electronic communications and transactionsacross society? That is, what is the monetary value of keeping PI protected, orhow much would it cost for others to have or use PI? Should the governmentregulate this information or is it better left to privacy industry or theindividual? Are these information and privacy issues merely personal decisionsthat individuals must evaluate to make their own choices and provide their own protection?
Thereare several things to consider when evaluating the cost of privacy. First, isdata sharing a public good? For example, Center for Disease Control may use thedata to trace the spread of disease in order to prevent further outbreak. Otherexamples include managing at risk populations, such as children under 16,people at risk of suicide, and the elderly. Moreover, consider groups ofextremists who seek to hide their activities. Should their data be trackable bythe government for national security concerns? Consider a person’s browser,phone system, and internet feed with their personalized advertisements; howmuch is this customization worth? Overall, when evaluating cost of privacy weneed to consider all of these tradeoffs. What is the potential gain fromkeeping data private and what is lost by doing so?
Asa policy analysis team for a national decision maker, your team’s tasks are:
Task1: Develop a price point for protecting one’s privacy and PI in variousapplications. To evaluate this, you may want to categorize individuals intosubgroups with reasonably similar levels of risk or into related domains of thedata. What are the set of parameters and measures that would need to beconsidered to accurately model risk to account for both 1) characteristics ofthe individuals, and 2) characteristics of the specific domain of information?
Task2: Given the set of parameters and measures from Task 1, model for cost ofprivacy across at least three domains (social media, financial transactions,and health/medical records). In your base model consider how the tradeoffs andrisks of keeping data protected affect your model.You may consider giving someof the tradeoffs and risks more weight than others as well as stratifyingweights by subgroup or category. Consider how different basic elements of thedata(e.g. name, date of birth, gender, social security or citizenship number)contribute to your model.Are some of these elements worth more than others? Forexample, what is the value of a name alone compared with value of