Recommandé
Recommandé
Contenu connexe
Tendances
Tendances (20)
En vedette
En vedette (19)
Similaire à Privacy Preserving Data Mining
Similaire à Privacy Preserving Data Mining (20)
Dernier
Dernier (20)
Privacy Preserving Data Mining
- 1. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References PRIVACY PRESERVING DATA MINING(PPDM) Romalee Amolic AISSMS College of Engineering Department of Computer Engineering March 23, 2017
- 2. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Outline 1 Introduction 2 Literature Survey 3 Methodology Used 4 Algorithms used 5 Advantages and Disadvantages 6 Conclusion 7 Future Scope 8 References
- 3. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Introduction SOCIAL ISSUE: Big data is mined so as to find out interesting patterns.Every person involved,is concerned about the leakage of private data i.e privacy of the individual’s data. MOTIVATION: Today privacy of data is one of the most serious concerns which people face on an individual as well as organisational level and it has to be dealt with in an effective manner.Hence,I chose this topic as privacy is the most important need of the hour.
- 4. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Introduction PROBLEM STATEMENT: We should make a comparative study of the available data mining and preserving techniques so as to learn and derive better methods to preservethe privacy of data. OBJECTIVES AND OUTCOMES: By using PPDM we can achieve the objectives of preserving the privacy of data as per the requirements of the user.By using various algorithms we can mine important rules from the database while still preserving the privacy of the data.
- 5. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Introduction of Proposed System Data mining takes place at various levels. The three entities can be categorisedas : Data Provider is the one who provides the data. Concern:Whether he can control the sensitivity of the data he provides to others. Data collector is the user who collects data from data providers and then publish the data to the data miner. Concern:To guarantee that the modified data contain no sensitive information but still preserve high utility. Data Miner is the user who performs data mining tasks on the data.Concern:How to prevent sensitive information from appearing in the mining results
- 6. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Literature Survey: The randomization method: The randomization method is a technique for privacy-preserving data mining in which noise is added to the data in order to mask the attribute values of records. The k-anonymity model and l-diversity: In the k-anonymity method, we reduce the granularity of data representation with the use of techniques such as generalization and suppression.[4] Association rule mining can probe to be the best method to preserve the privacy[2].This is one of the latest technologies and methods.It tries to eliminate the flaws if any in the previous methods. Based on in-depth study of the existing data mining and association rule mining algorithms, a new mining algorithm of weighted association rules can be proposed.It greatly reduces the time of input and output, and improves the efficiency of data mining.
- 7. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References In addition to that fuzzy association rules have been developed so as alter the support and confidence of rules as per the requirements. Data mining can be done at various stages[3].This paper tries to explore various PPDM techniques based on proposed PPDM classification hierarchy. The categorization of data mining can be done into: a>Centralized and b>Distributed data mining In addition to the normal methods of anonymization and associative rule mining the methods of perturbation and cryptography are discussed in detail.[3] In order to deal with these issues there might be balance between the privacy and utility of the data.This is the most important reason for the large research and development in this field.[1]
- 8. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Work Significance randomization Noise is added to the data k-anonymity Prevention of record linkage l-diversity Prevention of record and attribute linkage Association rule mining Finding out sensitive rules Fuzzy association rules Alter the support and confidence of rules. Perturbation Individuals can choose privacy level Cryptography Hiding information using cryptography.
- 9. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA PROVIDER Refuse to provide such data.
- 10. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA PROVIDER Refuse to provide such data. Negotiate with the data collector
- 11. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA PROVIDER Refuse to provide such data. Negotiate with the data collector Limit the access
- 12. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA PROVIDER Refuse to provide such data. Negotiate with the data collector Limit the access HOW?
- 13. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References How to limit the access? LIMIT ACCESS Anti- tracking extensions Disconnect Do Not Track Me Ghostery Do Not Track (DNT) Encryption tools MailCloakTorChat Advertisement and script blockers AdBlock Plus FlashBlock NoScript
- 14. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR The methods of PPDM are used for preserving the data of the data collector.We shall consider k-anonymity. Four types of attributes: *Identifier (ID) *Quasi-identifier (QID) *Sensitive Attribute (SA) *Non-sensitive Attribute (NSA)
- 15. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR The methods of PPDM are used for preserving the data of the data collector.We shall consider k-anonymity. Four types of attributes: *Identifier (ID) *Quasi-identifier (QID) *Sensitive Attribute (SA) *Non-sensitive Attribute (NSA)
- 16. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR The methods of PPDM are used for preserving the data of the data collector.We shall consider k-anonymity. Four types of attributes: *Identifier (ID) *Quasi-identifier (QID) *Sensitive Attribute (SA) *Non-sensitive Attribute (NSA)
- 17. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR The methods of PPDM are used for preserving the data of the data collector.We shall consider k-anonymity. Four types of attributes: *Identifier (ID) *Quasi-identifier (QID) *Sensitive Attribute (SA) *Non-sensitive Attribute (NSA)
- 18. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR The methods of PPDM are used for preserving the data of the data collector.We shall consider k-anonymity. Four types of attributes: *Identifier (ID) *Quasi-identifier (QID) *Sensitive Attribute (SA) *Non-sensitive Attribute (NSA)
- 19. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA COLLECTOR k-anonymity: Every tuple in the anonymized table is indistinguishable from at least k-1 other tuples along the quasi-identifiers k-anonymity can be implemented as: a.Generalization. b.Suppression.
- 21. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References DATA MINER The data miner makes use of the fuzzy association rule mining for data preservation. Association rule mining aims at finding interesting associations among large sets of data items. Steps involve: 1]Finding frequent item sets 2]Generate strong association rules Thereafter the fuzzy association rule mining technique can be used to alterthe level of a)Support and b)Confidence
- 22. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Fuzzy association rule mining Terms used: R->Interesting rules Rh->Sensitive rules D->Dataset MST->Mininum Support Threshhold MCT->Minimum Confidence Threshold
- 23. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 24. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 25. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 26. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 27. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 28. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 29. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 30. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Steps can be summarised as follows: Step 1:Fuzzification Step 2:Fuzzy Inferencing Step 3:Aggregation of all outputs Step 4:Defuzzification A association rule is of the form: x → y Association rule Antecedent Consequence
- 31. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Step 1:Fuzzification Fuzzy statements in the antecedent are resolved to a degree of membership between 0 and 1.
- 32. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Step 2.Fuzzy Inferencing(Implication Method): Truth value for the antecedent of each rule is computed and applied to the conclusion part of each rule.Degree of support is used. If the antecedent is only partially true, (i.e., is assigned a value less than 1), then the output fuzzy set is truncated according to the implication method. If the consequent of a rule has multiple parts, then all consequents are affected equally by the result of the antecedent. min: truncates the consequent’s membership function
- 34. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Step 3.Aggregation: It is the process where the outputs of each rule are combined into a single fuzzy set. The input? List of truncated output functions returned by the implication process for each rule. The output of the aggregation process is one fuzzy set for each rule. Here, all fuzzy sets assigned to each output variable are combined together to form a single fuzzy set.
- 36. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Step 4.Deffuzzification: In Defuzzification, the fuzzy output set is converted to a crisp number. Some commonly used techniques are the centroid and maximum methods.
- 38. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Advantages and Disadvantages ADVANTAGES: Protection of personal information. Development of KDD. Sharing of sensitive data for analysis. DISADVANTAGES: Balance between utility of the data and privacy. Loss of data. Availability of personal data.[7]
- 39. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Advantages and Disadvantages ADVANTAGES: Protection of personal information. Development of KDD. Sharing of sensitive data for analysis. DISADVANTAGES: Balance between utility of the data and privacy. Loss of data. Availability of personal data.[7]
- 40. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Advantages and Disadvantages ADVANTAGES: Protection of personal information. Development of KDD. Sharing of sensitive data for analysis. DISADVANTAGES: Balance between utility of the data and privacy. Loss of data. Availability of personal data.[7]
- 41. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Conclusion PPDM techniques can be used to make the data private and secure.These ensure efficient privacy preserving of data. The use of existing algorithms works towards the direction to reduce the impact of PPDM on the source database.
- 42. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References Future Scope Since, no such technique exists which overcomes all privacy issues, research in this direction can make significant contributions. The study can be carried out using any one of the existing techniques or using a combination of these or by developing entirely a new technique. Since, no such technique exists which overcomes all privacy issues, research in this direction can make significant contributions. The study can be carried out using any one of the existing techniques or using a combination of these or by developing entirely a new technique. Convex optimization method can be extended for any kind of association rules.
- 43. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References References [1] LEI XU, CHUNXIAO JIANG, (Member, IEEE), JIAN WANG, (Member, IEEE), JIAN YUAN, (Member, IEEE), AND YONG REN, (Member, IEEE) ”Information Security in Big Data:Privacy and Data Mining” [2] Lei Chen,’The Research of Data Mining Algorithm Based on Association Rules’ [3] Jisha Jose Panackal1 and Dr Anitha S Pillai2 Associate Professor Department of Computer Applications, Vidya Academy of Science and Technology, Thrissur, Kerala, India ’Privacy Preserving Data Mining: An Extensive Survey’ [4] Charu C. Aggarwal IBM T. J. Watson Research Center Hawthorne, NY 10532 Philip S. Yu University of Illinois at Chicago ’A General Survey of Privacy-Preserving Data Mining:Models and Algorithms’ [5] D. Jain, P. Khatri, R. Soni, and B. K. Chaurasia, ’Hiding sensitive association rules without altering the support of sensitive item’.
- 44. PPDM Romalee Amolic Introduction Literature Survey Methodology Used Algorithms used Advantages and Disad- vantages Conclusion Future Scope References [6] J.-M. Zhu, N. Zhang, and Z.-Y. Li, ’A new privacy preserving association rule mining algorithm based on hybrid partial hiding strategy’ Cybern. Inf. Technol., vol. 13, pp. 4150, Dec. 2013. [7] Privacy Preserving Quantitative Association Rule Mining Using Convex Optimization Technique’ [8] International Journal of Advanced Research in Computer and Communication Engineering Vol. 2, Issue 4, April 2013 Copyright to IJARCCE ww.ijarcce.com 1677 ’Privacy Preserving Data Mining’ Seema Kedar, Sneha Dhawale, Wankhade Vaibhav [9] http://donottrack.us/ [10] http://webpages.uncc.edu/xwu/career/ [11] http://www.intechopen.com/