Insiders are considered as the weakest link. The digital records of a person's Facebook likes against motivational quotes can be used for automatic and accurate prediction of sensitive attributes related to their personality traits depression, and their views against company/government policies, etc. Such analysis will help organization to take proactive measures against vulnerable insiders. Insiders managing their impressions differently than their basic personality traits can also be identified. Deep learning models can be utilized to learn and map the association among extracted features and insider behavioral patterns. Further, reinforcement techniques can be used to select appropriate motivational quotes in order to collect additional data required for further analysis. At the same time, the same exposed motivational messages on insider's social platform can aid to improve their psychological health over a time. However, due to implications involved in data collections related to personalization and data collection privacy, the authors have quoted their work in terms of this concept chapter only.
TopIntroduction
Many commercial applications exist in current digital markets that try to predict personal information to improve their products or services by invasions of user’s privacy. Individual traits can also be predicted from written texts or answers to a psychometric test. Similarly, employee behaviors on social media like Facebook, Instagram or Twitter can be analyzed; although such analysis presents serious challenges considering data privacy norms (Michal, David, & Thore, 2013). David Stillwell in 2007 (Michal, Tadesse, Bo, & Liang, 2018) created myPersonality, a Facebook App in his psychological research. In 2013, publicly available Facebook Page Likes of all users restricted by Facebook by implementing strict privacy policies. However, organizations; especially critical government agencies which are responsible for tackling national security issues; can send motivational messages or quotes daily to their employees to collect their likes and analyze their psychological behavior in order to predict any insider related threats in advance. Here, an insider can be any employee who has access to the critical information systems of an organization and can pose threats for organizational security (Miltiadis, Alexios, Nikos, Marianthi, & Dimitris, 2010, & Sunita, 2018).
Several insider threat prediction models have been proposed in the literature (Michal, Tadesse, Bo, & Liang, 2018). The authors in (Miltiadis, Alexios, Nikos, Marianthi, & Dimitris, 2010) captured the user’s technological trait (using IDS, honeypot etc.) augmented with data from psychometric tests in order to analyze insider’s tendency to malicious acts and the stress level. The authors categorized users in Novice, Advanced, Administrator categories based on their access level and also in Low, Medium, High user sophistication/computer skill levels. Authors in (Michal, Tadesse, Bo, & Liang, 2018) tried to predict personality traits of Facebook users using the myPersonality project dataset, where they compared performance of different machine learning models and found XGBoost classifier outperforms, with prediction accuracy of 74.2%. Authors in (Park,Youngin,& Kyungho, 2018). utilized machine learning algorithms to find the possible malicious insider using Sentiment-140 dataset and found Decision Tree and K-Means models provided highest accuracy. Authors in (Cristina, et al. 2017) tried to predict personality traits from Facebook profile pictures. In their research, they found that extroverts/agreeable individuals use warm colored pictures with many faces that showcase their socializing nature; while neurotic ones use pictures of indoor places and production of selfies associated to narcissism. Authors in (Marco, et al. 2013). utilized PsychoFlickr dataset, based on ProFlickr users for identification of personality traits and validated against user’s self-assessed answers and ratings given by 12 independent assessors.
All above works have not considered impact of likes/comments against motivational messages available on social media for insider personality prediction. As insiders personality traits influences on their choice of Facebook profile pictures (Cristina, et al. 2017); similarly likes/comments to motivational messages can be based on the many personality related factors like person’s thoughts/views towards life/job/colleagues, current stress level, what kind of problems currently he is facing, what he thinks about existing challenging job conditions, any family problems if he is facing currently, his financial conditions etc. The analyses carried out in our work are aimed to examine following research questions:
- (1)
Research Question 1 (RQ1): Do comments/likes against motivational messages helps to predict insider’s personality?
- (2)
Research Question 2 (RQ2): Do extracted deep learning features from comments/likes against motivational messages can provide valuable information related to advanced threat prediction?
- (3)
Research Question 3 (RQ3): Do the same platform can be used for improving positive useful personality traits to improve employees performance during course of period?
The paper is organized as follows. In section II, Overview of the Personality Traits is given. In section III, we explain various categories of motivational quotes and we try to discover relevance of them in analyzing personality traits/stress conditions. In section IV, we propose the deep learning model that can be used for insider feature classification and prediction followed by conclusions in section V.