Provisioning privacy in collaborative filtering recommender system

A Recommender System (RS) is the most significant technique that handle the information overload problem of Retrieval Information by suggesting users with correct and related items. Today, abundant recommender systems have been developed for different fields and search on collaborative filtering (CF

Project Title

Provisioning privacy in collaborative filtering recommender system

Project Area of Specialization

Shared Economy

Project Summary

A Recommender System (RS) is the most significant technique that handle the information overload problem of Retrieval Information by suggesting users with correct and related items. Today, abundant recommender systems have been developed for different fields and search on collaborative filtering (CF) recommender system. There are several problems in the recommender system such as Cold Start, Synonymy, Shilling Attacks, Privacy, Limited Content Analysis and Overspecialization, Grey Sheep, Sparsity, Scalability and Latency Problem. The current research explored the privacy in CF recommender system, and defined the perspective privacy attributes (user’s identity, password, address, and postcode/location) which are required to be addressed. Using the base models as Homomorphic and Hash Encryption scheme, we have proposed a hybrid model Homomorphic Hash Encryption (H2E) model that addressed the privacy issues according to defined objectives in current study. Furthermore, in order to evaluate the privacy level, H2E was implemented in medicine recommender system and compared the consequences with existing state-of-the-art privacy protection mechanisms. It was observed that, H2E outperformed to other models with respect to determined privacy objectives. User profile is totally secure so that no one unauthorized person can get access to your profile. H2E provide best user authentication in order to achieve privacy of users profile and all attributes under their profile. Leading to user’s privacy, H2E can be a considered a promising model for CF recommender systems and its results better than the previous researches as the execution time of recommendation is far better than others which show that no one unauthorized get access to user’s personal information during recommendation generation or giving rating to specific item.

Project Objectives

Provisioning privacy in collaborative filtering recommender system

•           Find privacy factors:

There are many factors that need privacy. First we find out the factors that need privacy in collaborative filtering recommender system. So that user’s personal information is not reveal.

•           Determine privacy:

After finding the factors we determine how to provide privacy in collaborative filtering recommender system. For this purpose, we analyze all privacy provisioning techniques in collaborative filtering RS and select the best technique. After that we propose our new technique that achieve our privacy provisioning goal in collaborative filtering RS. At the end, compare results with the previous researches.

Project Implementation Method

Phase 1: Survey of related work and background

In this phase, we survey work related to our research, point out the many contributions of previous researchers by comprehensive knowledge about the history or background and all the work which is done so far has to achieve by studying the research papers and Journals or by views that are collected from an organization. A large amount of work has been done in the area. In the first phase, we present a brief but essential background on error control. We then proceed to describe the problems encountered when we apply collaborative filtering recommender system [31].

Phase 2: Analysis

Analysis phase involve identifying common patterns within the responses and critically analyzing them in order to achieve research aims and objectives. It involves the critical analysis of collected information to check genuineness and relevance with the required information.

Phase 3: Implementation

This phase involve the implementation of new approach by using some algorithms, tools, techniques or model.

• Data collection
• Database design
• Data preprocessing
• Registration of user
• Recommendation implementation
• Privacy implementation
• Front end design
• Testing
• Documentation

Phase 4: Comparative study and results

This phase present the comparison of previous approaches with the new approach which is implemented. Results are compared to see the difference between already existing and new approach so that we know what new and better measures new approach is providing to resolve privacy issue which previous approaches lack.

Phase 5: Thesis writing

This phase has to be conducted to write down all about research work. If comparative study goes successful then thesis writing starts with advantages of new approach and challenges so that others can accomplish them in future.

Benefits of the Project

The major benefit of our research is that user’s personal information will never be leaked out to third party. Their all data must be secure so the service provider and their internal employees’ not miss use the data

The second min benefit of our research is to proposed new privacy approach for securing user’s personal information.

Technical Details of Final Deliverable

A new model of privacy-preserving collaborative filtering recommender system, by using the base models homomorphic technique and Hash function encryption technique proposed a Hybrid model homomorphic hash encryption (H2E) , which allows the computations required for recommendations in a dispersed manner and preserves user privacy without compromising recommendation accuracy and efficiency. By using the hybrid model we achieve privacy to all the factors that are the main concern of users.

Mobile application of collaborative filtering recommender system that recommend medicines according to user’s previous rating.

Final Deliverable of the Project

Software System

Type of Industry

IT

Technologies

Shared Economy, Wearables and Implantables

Sustainable Development Goals

Responsible Consumption and Production

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com