Giannis Haralabopoulos

Giannis joined Horizon as Research Fellow a little over 6 months ago. He told us…

“On a Tuesday morning, 14^th of January 1997, I received my first computer. It was a Pentium II 300Mhz with 32MB of RAM. Not a high-end computer by that era’s standards, but more than enough to spark my interest towards computers. It took more than a decade to transform my interest to a subject of studies.

Twelve years later I (finally) was given the opportunity to attend a Master of Science in Computer Science. Prior to that, a four-year Bachelor of Science in Mathematics was attained, where I developed skills closely related to Computer Sciences, such as analytical thinking and structured reasoning. Towards the end of my M.Sc. I was given the opportunity to start a Ph.D. around Online Social Networks.

Growing older and studying social interactions online, I became interested in human nature as depicted by their online behaviour. My Ph.D. dissertation was titled “Online social networks: real-time graph sampling, multilayered information diffusion, and equality issues”, but the equality issues part was a simple mapping of the online plain at that time. Joining the Horizon research group gives me the opportunity to thoroughly study interdisciplinary subjects, including Human Computer Interactions and Social Networks.

Before joining Horizon, I worked in University of Strathclyde and University of Southampton. At Strathclyde I became familiar with Knowledge Exchange Partnerships and data driven projects, and at Southampton I was involved with multiple EU2020 projects and student supervision. Horizon gives me the opportunity to collaborate with research scientists from different fields in a research oriented environment.

As a new member of Horizon, I am trying to create a network of collaborations and disseminate our research. Our most recent paper “A multivalued emotion lexicon created and evaluated by the crowd” was presented in Social Networks Analysis, Management and Security 2018 and deals with the creation of pure emotion lexicons for machine learning purposes. In addition, our paper “Paid crowdsourcing, low income workers, and subjectivity” will be submitted in Mining Humanistic Data Workshop 2019 and presents our early findings with regards to human contributors and subjective tasks. We aim to present our machine learning contributions in a separate journal publication in 2019.

Going forward we aim to study human computer interactions from a health perspective, and secure external funding for these health related projects. One of our proposed future projects will involve analysis and processing of user generated data, with respect to the user privacy, for mental health detection purposes. A project that we aim to extend for medical transcripts and records.

Our goal is to improve well-being and produce non-intrusive human-centered studies and applications.”