m.e project presentation
TRANSCRIPT
Preserving User’s Privacy in
Personalized Search
Guide: Ms.S.L.Jany Shabu Name: Manoj Kumar.K
Reg No: 3331034 M.E - CSE
2
Objectives
• Personalized web search(PWS) is a search technique for providing better search results & viewing user history.
• Privacy issues are raised by revealing user’s private life.
• Aim to disclose the user’s identity in database during the PWS.
• Unique(Pseudo) identity or Group identity aims to recovery the privacy issues.
Literature SurveyS.No Title Author Journal/
ConferenceInference Limitations
1 Supporting Privacy Protection in Personalized Web Search
L.Shou, H.Bai, K.chen, G.Chen
IEEE-2014 Allows for customization of privacy needs.
Online decision towhether personalize or not.
User’s identity to server is exposed.
Effectiveness of personalization is reduced.
2 Online Anonymity for Personalized Web Services
Y. Xu, K. Wang, G. Yang & A.W.-C. Fu
ACM-2009 Anonymizer algorithm on user profiles by generating anonymous identification.
No personal information.
Delay in real-time response due to higher communication cost.
Higher Encryption/Decryption cost.
3 Efficient Privacy Protection in Personalized Web Search
G. Chen, H. Bai,L. Shou, K. Chen, & Y. Gao
ACM-2011 Greedy Utility algorithm for generalizing user queries in PWS.
Identification is required for server.
It requires more computational cost & recomputation of all queries.
4 Anonymizing User Profiles for Personalized Web Search
Y.Zhu, L.Xiong,C.Verdery
WWW Conference-2010
No identity privacy protection using Bayes-optimal privacy notion.
Online anonymity on user profiles by generating a anonymous id using anonymous proxy server.
Personalized Search results are not effective.
It is not portable for all users.
5 A Utility-Theoretic Approach to Privacy in Online Services
A. Krause E. Horvitz
AI Research – 2010
Probabilistic model used for predefined frequent queries.
Suggested to compromise privacy, if there is better search results.
For distinct queries, it is not applicable.
Privacy can’t be compromised, balancing both metrics is better.
5
Existing System
• Profile-based PWS do not support runtime profiling.
• User queries stored in database with entire identity of the particular user.
• Sensitive topics are detected using an absolute metric called surprisal.
• Entire expose of user profile to the database administrator.
6
Drawbacks
• Exposure of user queries with identity.
• Man-in-the-middle attack, invading the server & capturing user data.
• Server-side personalization.
• Revealing the user’s private life.
• Level of privacy protection is not achieved.
7
Proposed System
• For avoiding the exposure of user identity in existing system, the keywords in database are mapped by unique identity for each user & group of users.
• Supports runtime profiling for generalization of user queries.
• While viewing user histories by admin, it is linked with that particular user identity without revealing user profile.
• User Profile Encoding, Random Unique Identity generator & GreedyIL algorithms are used for disclosing user’s identity & generating unique identity for each user.
8
Advantages
• Enhances the stability of search by runtime profiling.
• Avoids exposure of user identity to database.
• Eavesdropping privacy attack can be prevented.
• Level of privacy protection is obtained in PWS.
• Without compromising both personalization & privacy.
9
Architecture Diagram
10
Modules
• User Registration New user is registering after validating entire profile & after creating account, an alphanumeric identity is assigned to that user & entire profile is encoded.• User Login After user login, there is search page, search results & view user history. While Searching, the visited pages are viewed in user history but from server side there is no identity of profile except unique identity who visited those pages.• Database Admin Admin can view registered user’s profile which is fully encrypted form mapped with pseudo identity. While viewing records of user history also, admin can view all search keywords & visited pages but it is mapped with unique identity. User can use all resources without bothering about privacy issues.
11
System Requirements
• Hardware Ram : 256MB & above Hard disk : 20GB & above
• Software Front end : JSP Database : MySQL
12
Screenshots of ModulesHome page
13
User Registration
14
15
16
Database Admin View
17
User Login
18
User Index Page
19
Search Page
20
Search Results Page
21
Viewing User History
22
User History Database View
23
Administrator Login
24
Admin Index Page
25
View of Encrypted Users
26
Publication Details
• Paper titled as “Preserving User’s Privacy in Personalized Search” is accepted & published in International Journal of Applied Engineering Research(IJAER) ISSN 0973-4562 Volume 9, Number 22 (2014) pp. 16269-16276.
http://www.ripublication.com/Volume/ijaerv9n22.htm
27
References
1. Lidan Shou, He Bai, Ke Chen, and Gang Chen, “Supporting Privacy Protection in Personalized Web Search,” IEEE Transactions on Knowledge and data engineering Vol:26 No:2, Year 2014.2. G. Chen, H. Bai, L. Shou, K. Chen, and Y. Gao, “Ups: Efficient Privacy Protection in Personalized Web Search,” Proc. 34th Int’l ACM SIGIR Conf. Research and Development in Information, pp. 615-624, 2011.3. Y. Zhu, L. Xiong, and C. Verdery, “Anonymizing User Profiles for Personalized Web Search,” Proc. 19th Int’l Conf. World Wide Web (WWW), pp. 1225-1226, 2010.4. A. Krause and E. Horvitz, “A Utility-Theoretic Approach to Privacy in Online Services,” Artificial Intelligence Research, vol. 39, pp. 633-662, 2010.5. K. Hafner, Researchers Yearn to Use AOL Logs, but They Hesitate, New York Times, Aug. 2006.
28
Contd…
6. J. Castelli-Roca, A. Viejo, and J. Herrera-Joancomartı´, “Preserving User’s Privacy in Web Search Engines,” Computer Comm., vol. 32, no. 13/14, pp. 1541-1551, 2009.7. J.Teevan, S.T. Dumais, and D.J. Liebling, “To Personalize or Not to Personalize: Modeling Queries with Variation in User Intent,” Proc. 31st Ann. Int’l ACM SIGIR Conf. Research and Development in Information Retrieval (SIGIR), pp. 163-170, 2008.8. K. Ramanathan, J. Giraudi, and A. Gupta, “Creating Hierarchical User Profiles Using Wikipedia,” HP Labs, 2008
29
Thank You