Conference Papers

Permanent URI for this collectionhttps://idr.nitk.ac.in/handle/123456789/28506

Browse

Author: search.filters.author.Chandrasekaran, K.Subject: search.filters.subject.Privacy

Search Results

Now showing 1 - 3 of 3
  • No Thumbnail Available
    Item
    A Privacy Preserved Data Mining Approach Based on k-Partite Graph Theory
    (Elsevier, 2015) Bhat, T.P.; Karthik, C.; Chandrasekaran, K.
    Traditional approaches to data mining may perform well on extraction of information necessary to build a classification rule useful for further categorisation in supervised classification learning problems. However most of the approaches require fail to hide the identity of the subject to whom the data pertains to, and this can cause a big privacy breach. This document addresses this issue by the use of a graph theoretical approach based on k-partitioning of graphs, which paves way to creation of a complex decision tree classifier, organised in a prioritised hierarchy. Experimental results and analytical treatment to justify the correctness of the approach are also included. © 2015 The Authors.
  • No Thumbnail Available
    Item
    Generating Privacy-Preserved Recommendation Using Homomorphic Authenticated Encryption
    (Institute of Electrical and Electronics Engineers Inc., 2017) Shanu, P.K.; Chandrasekaran, K.
    Online service providers started to providepersonalized recommendation to the users by collecting userprivate sensitive data. Traditionally the user private data isencrypted using a symmetric encryption algorithm beforestoring it in the cloud to provide another layer of security fordata at rest. It makes users' data secure from third parties, butnot the service provider. We propose a method that generatesrecommendations using homomorphically encrypted data in aprivacy preserved manner to provide protection against serviceprovider. We also verify the correctness of computations doneby a third parties and the service provider over encrypteddata using homomorphic authenticators and some securecryptographic protocols. © 2016 IEEE.
  • No Thumbnail Available
    Item
    Towards a Federated Learning Approach for NLP Applications
    (Springer Science and Business Media Deutschland GmbH, 2021) Prabhu, O.S.; Gupta, P.K.; Shashank, P.; Chandrasekaran, K.; Divakarla, D.
    Traditional machine learning involves the collection of training data to a centralized location. This collected data is prone to misuse and data breach. Federated learning is a promising solution for reducing the possibility of misusing sensitive user data in machine learning systems. In recent years, there has been an increase in the adoption of federated learning in healthcare applications. On the other hand, personal data such as text messages and emails also contain highly sensitive data, typically used in natural language processing (NLP) applications. In this paper, we investigate the adoption of federated learning approach in the domain of NLP requiring sensitive data. For this purpose, we have developed a federated learning infrastructure that performs training on remote devices without the need to share data. We demonstrate the usability of this infrastructure for NLP by focusing on sentiment analysis. The results show that the federated learning approach trained a model with comparable test accuracy to the centralized approach. Therefore, federated learning is a viable alternative for developing NLP models to preserve the privacy of data. © 2021, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.