Research

MY Research Topics

Discrimination in Data Mining

Discrimination is treating people unequally according to their membership in a specific group, class, or category. Membership criteria can encompass race, gender, native country, religion, age, etc. A remarkable amount of legal regulations ban discriminatory decisions on an individual basis. Although the legal regulations against the discrimination of individuals are clear, recent cases in the EU and in the US show that there are still effective direct and indirect discrimination in professional life and in provided services. In this project, we are analyzing the causes of discrimination in the historical data and proposing methods of historical discrimination prevention in the decision making.

Publications

Classification Using Private Data

Corporations are retaining ever-larger corpuses of personal data; the frequency or breaches and corresponding privacy impact have been rising accordingly. One way to mitigate this risk is through use of anonymized data, limiting the exposure of individual data to only where it is absolutely needed. This would seem particularly appropriate for data mining, where the goal is generalizable knowledge rather than data on specific individuals. In practice, corporate data miners often insist on original data, for fear that they might "miss something" with anonymized or differentially private approaches. In this project, we are providing a theoretical justification for the use of anonymized data in a particular data mining task "classification".

Publications

Personality Recognition Using Deep Learning

Personality recognition is a critical task for real life applications such as targeted marketing. In psychology, an individual's personality is identified by the "Big Five" traits: 1) Extraversion vs. Introversion 2) Emotional Stability vs. Neuroticism 3) Agreeableness vs. Disagreeable 4) Conscientiousness vs. Unconscientious 5) Openness to experience. In this project, we are extracting an individual's linguistic cues to predict his/her personality in the Big Five traits.

Publications

EXPERIENCE

MY Educational and Professional Timeline

  • 2014 - Present (Research)
  • 2014 - Present (Teaching)
  • Summer 2014
  • 2013 - 2014
  • Summer 2013
  • 2010 - 2013
  • 2006 - 2010
  • Purdue University, West Lafayette, IN, USA


    Research Assistant

    08/2014 - Current

    PhD in Computer Science

    Thesis: Classification Using Private Data

    Project 1: Privacy Preserving Data Mining in the Cloud, Purdue Research Foundation. Authored grant proposal that was approved

    Project 2: Mining Anonymized Data, Cybersecurity Consortium of Northrop Grumman Corporation. Authored grant proposal that was approved

  • Purdue University, West Lafayette, IN, USA


    Effective Teaching in CS

    08/2015 - 10/2015

    Passed the CS department teaching certification.


    Teaching Assistant

    08/2015 - 12/2015

    Lab Instructor in the Problem Solving and Object-Oriented Programming (CS 180) course

    Supervisor: Lorenzo Martino


    Graduate Student Board Mentorship Program

    10/2016 - 12/2016

    Guiding first year PhD students for course work and research


  • Software Engineer Intern

    05/2014 - 08/2014

    Asymmetric/Symmetric Encryption Implementation for a Vehicle Infotainment System

  • Purdue University, West Lafayette, IN, USA


    Research Assistant

    08/2013 - 05/2014

    Master of Science in Computer Science

    Project: Managing Private Data in the Cloud, Qatar Foundation


  • Visiting Scholar

    05/2013 - 07/2013

    Collaborator: Ryan Riley

    Project: Managing Private Data in the Cloud, Qatar Foundation

  • Purdue University, West Lafayette, IN, USA


    Research Assistant

    11/2011 - 05/2013

    Master of Science in Computer Science

    Project: A Framework for Managing the Assured Information Sharing Lifecycle, National Science Foundation

  • 09/2006 - 07/2010

    Bachelor of Science in Computer Engineering (Diplôme de Génie Informatique)

    Thesis: Automatic Annotation of Web Service Descriptions (Annotation Automatique des Description de Services Web)

    Advisor: Vincent Labatut

What I have achieved

MY Publications

Conferences and Workshops

  1. Koray Mancuhan, and Chris Clifton. " Statistical Learning Theory Approach for Data Classification with l-diversity. ," 2017 SIAM International Conference on Data Mining (SDM) (to appear) , SIAM, 2017

  2. Koray Mancuhan, and Chris Clifton. " Decision Tree Classification on Outsourced Data. ," 20th ACM SIGKDD international conference on Knowledge discovery and data mining (KDD) Workshop on Data Ethics. , ACM, 2014

  3. Koray Mancuhan, and Chris Clifton. " Discriminatory decision policy aware classification. ," 2012 IEEE 12th International Conference on Data Mining (ICDM) Workshops. , IEEE, 2012

Journals and Archives

  1. Koray Mancuhan, and Chris Clifton. " K-Nearest Neighbour Classification Using Anatomized Data. ," arXiv preprint arXiv:1610.06048 (2016) , arXiv, 2016

  2. Koray Mancuhan, and Chris Clifton. " Combating discrimination using Bayesian networks. ," Artificial intelligence and law 22.2 (2014): 211-238 , Springer, 2014

MY Grants

  1. Chris Clifton (PI) and Koray Mancuhan,"Mining Anonymized Data", Northrop Grumman Corporation , 1/1/16-12/31/16, $94,941

  2. Chris Clifton (PI) and Koray Mancuhan,"Privacy Preserving Data Mining in the Cloud", Purdue Research Foundation , 08/1/14-8/1/15, $26,403