ASmartPrivacyModelofDynamicPricing资源-CSDN文库

需积分: 5 60 浏览量 2023-05-25 14:36:05 上传评论收藏 1.32MB PDF 举报

资源推荐

资源详情

资源评论

For office use only

Team Control Number

For office use only

88508

Problem Chosen

2018 ICM Summary Sheet

A Smart Privacy Commodity: a Quantitative Model of Dynamic Pricing Strategy

This article presents a model to precisely quantify the price set and risk of private information(PI), and with it we

can get the optimal strategy of the privacy management.

To precisely quantify the price set of PI and get the optimal strategy of the privacy management, we innovatively

decomposed our model into two sub-problems, they are model of PI price set and model of privacy risk.

First, we created the model of PI price set on individual scope, community scope and nation scope. At the first step,

we gathered the basic characteristics of individuals in the domain of social media, financial activity, medical care

and e-commerce, and defined the PI, PP, IP given in task 5. We weighted combination of entropy method and

AHP and decided the weight of the four domains. Then based on TOPSIS comprehensive evaluation method and

PI’s consistency tradeoffs, we built the model of privacy price set on individual scope. At the second step, to

precisely measure the price set of PI on community scope, we learnt from the multiway tree theory and divided

people by their age, occupation, income to make our sub-community have the similar characteristic. Considered the

requirement of task, we precisely quantified the risk perception (which is determined by privacy type,

generational differences, network effects, data breach and community difference). From generational differences

we can launch a conclusion that people in 20-30 years old are most willing to share their information, and the

tendency is decrease by the age turns old. The data breach will lead a heavy reducement of the willingness of

information share, and can hardly resurgence in short time. Considering the time factor, we combined the risk

perception and markov chain model, and got the dynamic alter process of the risk perception by the year. In the

calculation of the matrix of transition probability, we used the bayesian theorem. Then we combined the model of

privacy price set on individual scope and model of risk perception and got the price set of PI on community scope.

At the third step, we analyzed the price set model on nation scope from three aspects. They are market aspect,

policy aspect and culture aspect. On the market aspect, we used the bayesian nash equilibrium; On the policy

aspect, we considered the macroscopic readjustment and control policy, privacy authority policy and strike

information leak and illegal transaction policy; On the culture aspect, we considered the atmosphere of information

transparency. From the analysis we got the price set model on the nation scope, the amount of all four domains is

783$.

Second, for the risk model, we qualitatively analyzed the different type of risk which faced by different community

in different domains, and measured these risks. Then we answered the question given in task4, task6 and task7.

In addition, we used real data to examine our model. The inaccuracy of our model is no more than 0.17, which

means our model is highly practical. We also put our model through error analysis and robustness analysis. The

outcome shows our model has well robustness, which means our model is highly reliable. At last, we summarized

our advantage and weakness, and blueprinted further development.

The innovation of the article is that we consider enough factors. Besides, in the logistic model of network effects,

we randomly chose 2012 and 2032 year, then using visualization method to compare their variation, and answer the

question in task 6. We considered PI in different domains have different influence (good for public interest or harm

national security), and quantified the external effect, which made our model optimized. In conclusion, our model

answered all the given questions, and has its creativeness and practicability.

Keywords: Pricing Model of PI; Risk Model of PI, Topsis Method; Markov Chain; External Effect

更多数学建模资料请关注微店店铺“数学建模学习交流”

https://k.weidian.com/RHO6PSpA

February 13, 2017

Dr.Decision Maker

Director of ICM

Dear Dr. Decision Maker:

Thank you for working with us to design a plan to quantify the cost of privacy of electronic communications

and transactions across society. We have developed a model to address your concerns about the cost of

privacy.

Through the model, we found out the value and risk of PI in all fields. The value and risk can be divided into

two categories according to their nature: personality and economic rights. Therefore, we suggest that when

discussing the privacy rights of information, it should be divided into two categories. When PI plays the role

of safeguarding the value or function of human dignity, it should give the right of personality determination;

when PI plays the value or function of the property, it should be given the rights of property; when playing

dual functions, it should be based on double recognition.

In the model we evaluated different elements. The conclusion is that the importance of the different elements

of the data is various. For example, the value of ID number is higher than that of age. People are also

different in their willingness to expose them, and their losses after being illegally used are also quite

different. Therefore, our advice on the protection and use of information is as follows.

First, the information is divided into personal general information and personal sensitive information. On the

basis of the distinction, it strengthens the protection of sensitive personal information and strengthens the

use of general personal information. The classification is based on the PI types. Such as age or sex is a less-

sensitive message which can be enhanced use. More sensitive personal information, such as hobbies and

political preferences, should be based on respect for the individual's wishes. Bank card number, physical

condition and other the most sensitive information should be strictly protected and strictly controlled its use.

Reconcile the conflict of needs protection and utilization of personal information and achieve the balance of

interests.

Our generational difference model reflects intergenerational differences in the population. The 20-30 age

group has an open attitude toward the sharing of personal information, much higher than other age groups.

However, because young people have less social experience, they are not aware of the risks posed by the

sharing of personal information. Therefore, the blind sharing of personal information may bring greater risks.

The government should strengthen its information security education for this group. While other age groups

remain closed to information sharing, which will reduce the efficiency of using personal information. The

government should strengthen the protection of information and reduce their vigilance.

Through the model we also found that the use of personal information will have external benefits which

cannot be detected by general measures. This is of great importance when considering the value of the entire

social PI. This external effect may be positive or negative. For the positive externalities, such as the use of

PI can reduce the overall social losses in the field of disease control. These are the market regulation can not

capture and government support and regulation. Similarly, negative externalities also require government

control.

An information leak event is a contingency which will cost so much. As we can see from our risk perception

model, when Information leakage event occurs and exposed to public view, the public's trust in information

Team # 88508 Page

of 29

sharing will rapidly decline and it will be hard to recover in a short time. After the information leaked，the

probability of their loss is greatly increased resulting from harassment and fraud. This will affect the

personal life and social stability. These losses are caused by the mismanagement of the information

controller, which entitles the individual to claim against the enterprise or agency.

Our model defines PI values and risks that provide direction for the utilization and protection of PI. Among

these, the trade-off between the use of information and information protection is an eternal topic. Personal

information is "information that can identify one's identity," and identifiable factors can have an impact on

the protection of the rights and interests of the main body of information and the utilization of the company.

In the model of risk identification

，

we find that most of the risk comes from "Information that can identify

one's identity". But from the value model, it can be seen that "Information that can identify one's identity" is

not an important factor. Therefore, "personal factors eliminating" can be done during the collection, storage,

processing and utilization of PIs. For example, when personal information is used in targeted marketing of

financial transactions, removing a single person's explicit identification does not affect the analysis of the

characteristics of a consumer group. Therefore, in the collection process, unless the consent of the main

body of information ， the relationship between the personal characteristics and the specific identification

information should be cut off.

At the same time, focusing on information leakage is the most important aspect of information protection. In

the era of big data, Internet users in the personal information, comments, pictures, hobbies, transaction

information, visit the website, etc. were recorded by the enterprise. Coupled with powerful data mining

technology, making information interrelated, once the data leaked, personal information security is facing a

great threat.

In order to reduce the risk of leakage and reduce losses, the advice we give based on our model as follow.

The conclusion for in the country-level pricing model is that the national policy of combating the crime of

information theft will reduce the information leakage rate. Therefore, we suggest that the state should strictly

legislate to crack down on crimes. Second, raising security awareness from the individual, business, social

level, and making information safe to share. Third, prevent staff leaks is very important. Data breach

survey report in 2008 shows that only 18% of the information leakage is due to Internal staff leaked. But its

destructive power is obviously greater than the external damage. Therefore, we must do a good job of the

confidentiality of internal staff.

Sincerely,

Team#88508

Team # 88508 Page

of 29

A Smart Privacy Commodity:

A Quantitative Model of Dynamic Pricing Strategy

1 Introduction

1.1 Background................................................................................................................. 4

1.2 Restatement and Clariﬁcation.....................................................................................4

1.3 Overview of Our Work............................................................................................... 4

2 General Assumptions and Justifications...............................................................................5

3 Data Sources and Variable Description................................................................................ 5

3.1 Data Sources............................................................................................................... 5

3.2 Variable Description....................................................................................................6

4 Privacy Pricing Model.......................................................................................................... 7

4.1 The Perspective of Individual.....................................................................................7

4.1.1

Features and Parameters of the Selection.........................................................7

4.1.2

Wegihted Combination of Entropy Method and AHP......................................8

4.1.3

TOPSIS Comprehensive Evaluation Method...................................................9

4.1.4

PI’s Consistency Tradeoffs............................................................................. 10

4.1.5

Individual Comprehensive Pricing................................................................. 10

4.2 The Perspective of Group......................................................................................... 11

4.2.1

The Classification of Sub-groups Based on Multiway Tree.......................... 11

4.2.2

The Group Pricing Model Based on Markov Chain and Bayes Theorem..... 11

4.3 The Perspective of Entire Nation............................................................................. 14

4.3.1

Market Model Based on Game Theory.......................................................... 15

4.3.2

The Perspective of Policy............................................................................... 15

4.3.3

The Perspective of Culture............................................................................. 16

4.3.4

The Overall Results of National Pricing Model.............................................16

4.3.5

The Optimization of Pricing Model - Considering the External Effect.........17

5 Risk Model of PI.................................................................................................................18

5.1 The Definition of Risks of Each Group in All Fields.............................................. 18

5.2 Risk Measurement and Calculation..........................................................................19

6 Testing the model................................................................................................................ 19

6.1 Examination Using the Real Data............................................................................ 19

6.2 Error Analysis........................................................................................................... 20

6.3 Stability Test..............................................................................................................20

7 Strengths and Weaknesses.................................................................................................. 21

7.1 Strengths....................................................................................................................21

7.2 Weaknesses................................................................................................................21

7.3 Future Work...............................................................................................................21

8 Conclusion.......................................................................................................................... 21

References.................................................................................................................................21

Appendix...................................................................................................................................22

Team # 88508 Page

of 29

Introduction

1.1

Background

“App makers sometimes like to test the limits to see what they can get away with, before users decide to

turn off a feature because they aren’t comfortable with the amount of personal data being provided, or the

audience that it's being shared with.”

- Daniela Relph, Alan Turing Institute, quoted from a BBC article

“People always want to give yourself some privacy space, just like you will always be standing in front of

the shadow of you blocking the line of sight of the light.”

- Laura Sheeter, LSE, quoted from a BBC article

At the moment, webcasts and social media are sweeping across the globe. Life sharing and information

dissemination have become the main theme of today's society. Especially in today's Internet era and big data

era, every move of our lives is recorded and controlled. Social media has become a window to people's lives,

and strangers can know about it through this window. Some people believe privacy is very important and it

needs protection, while others think privacy can be shared. Therefore, the discussion of privacy is

particularly necessary. We view privacy as a commodity that can be quantified and traded and to establish

our model.

1.2

Restatement and Clariﬁcation

Our team is certain to construct a policy model, which include a set of policy recommendations that will

create an efficient privacy pricing plan. To ﬁnd out the optimal strategies, we need to construct models, run

simulations and present the visualized results. Our model should be scalable, multilayered and dynamic.

However, trade-offs should be made if some objectives contradict.

We aim to solve the following problems:

 Categorize individuals into various subgroups and develop a price point for protecting one's privacy and

PI in various applications. And deﬁne a set of parameters, measures from the perspective of individuals

and specific domain of information.

 Based on some trade-offs, a privacy cost model is built by grouping the weights of the four domains

(social media, financial transactions, health/medical records, electronic commerce).

 The privacy pricing model is set up from the perspective of individuals, groups and entire nations

respectively. The model needs to take into account the relationship between supply and demand and

personal preference.

 State the assumptions and constraints of the model established and define whether information privacy

is a fundamental human right. Introduce dynamic elements to explore changes in people's values.

 Answer the question of whether there are generational differences in perceptions of the risk-to-benefit

ratio of PI and data privacy. Explain the effects of age and distinguish the influence of PI, PP and IP.

 Using the optimized method to capture the network effect of data sharing, to answer the question of

which subjects are affected, and the responsibilities that come with the community.

 How to adjust the model if the privacy data is lost or abused? Answer the question of how the

responsible institution of the violation should perform the liability.

 In addition to using real data to test the stability of the model, sensitivity analysis is also used.

1.3

Overview of Our Work

We divided the PI into four areas: social media, financial transactions, health records, and e-commerce, then

analyze the properties and characteristics of PI ,and answer the problems raised by the task 5 that difference

between the PI, PP, and IP. At the same time integrated Task1-7, we set up pricing models from three levels:

individuals, groups and countries. We conducted a selection of individual characteristics and specific

information areas characteristics, and combine Analytic Hierarchy Process and Entropy method to determine

the weight of characteristics in the four areas. Finally, based on the TOPSIS method and combined the

Information Feature Completeness, We established a PI pricing model of individual level.

剩余28页未读，继续阅读

评论收藏

内容反馈

xiaoshun007～

粉丝: 3777
资源: 3146

A Smart Privacy Model of Dynamic Pricing

最新资源

A Smart Privacy Model of Dynamic Pricing

Theory of Asset Pricing

Theory of Asset Pricing by George Pennachi

A GARCH option pricing model with α-stable innovations

Theory of Asset Pricing--Pennacchi 资产定价理论课后习题答案

a five-factor asset pricing model.pdf

A fractional Merton model for option pricing

Answer book for Theory of Asset Pricing by George Pennachi

价值收益与优化-菲利普斯 pricing and revenue optimization

Option pricing formula for a new stock model.pdf

翻译 International tests of afive-factor asset pricing model.docx

A Mixed Brownian-Poisson-fractional Model for option pricing

the capital asset pricing model theory and evidence.pdf

边缘计算资源分配物联网环境

A novel method to update dynamic pricing scheme for Diff Serv (2005年)

asset pricing

an empirical investigation of the arbitrage pricing theory.pdf

linux的工具，适合源代码 pricing a thing with a lot of tricks 的编写

A Pricing Model for Big Personal Data

A Kind of Accelerated AOS Difference Schemes For Dual Currency Option Pricing Model

相关实用应用程序（Windows可用）

免费可用的ChatGPT网页版.zip

ChatGPT使用总结：150个ChatGPT提示词模板（完整版）

chromedriver-win64.zip

全国计算机二级WPSoffice精选350道选择题题库（含答案）.pdf

哈尔滨工业大学-ChatGPT调研报告-2023.3.6-94页.pdf

HAI-2024斯坦福AI指数报告（中文译版）.pdf

2023泛娱乐社交出海手册-ZEGO即构科技

4个亲测好用的ChatGPT4渠道

毕业设计的概要介绍与分析

最新资源