Sophia's journey to the DoSSIER project

User Rating: 5 / 5

Star ActiveStar ActiveStar ActiveStar ActiveStar Active

--   Interview with Sophia Althammer   --

S: Tell us a bit about your background!

S: I studied in my Bachelors and Masters Mathematics at the TU Munich, during my masters I then focused on Machine learning and data science. During an internship at a consultancy I first got introduced to the topic of Natural language processing and information extraction and was absolutely amazed by the research done in this field. It was clear to me that I wanted to write my master thesis about NLP and then I found my topic about representation learning for language in the patent domain, so my background is in Mathematics, Machine learning and NLP.

S: How did you find out about DoSSIER and what made you decide to apply?

S: I found the DoSSIER project on a website for positions and phds in nlp and immediately loved the projects focus of IR and NLP in the medical, legal and patent domain. Since I already had some experience in NLP in the patent domain and since I loved the idea of moving to Vienna, I decided to apply!

S: How long are you already working on the project?

S: I am working on the project since 9 months now, I started in July 2020.

S: What are your initial thoughts on DoSSIER and your experience?

S: I enjoy being part of a big research project and I enjoy the training activities and the exchange with the other PhD students. I would love to meet them all once in person and hopefully a summer/winter school in person is soon possible. I love how much I already developed as a researcher, how many new experiences I gained by getting an overview of the research and trends in IR, by reading inspiring papers, by doing experiments myself and by writing, publishing and presenting my own paper at ECIR 2021.

S: What project are you currently working on?

S: Currently I am working on some different tasks. Me and my colleague Arian were participating in the COLIEE competition 2021, where we applied our research methods for the task of legal case retrieval. Besides that I am interested in transferring advances of neural IR and NLP in web search to professional search in the medical, legal and patent domain and also study the cross-domain transferability. Here my interest lies especially in methods which are computationally feasible for a production-ready systems.

S: Tell us about your team environment. What do you enjoy about collaborating with your small team and the larger lab?

S: I work closely with some of the ESRs in the DoSSIER project on some competitions or evaluation campaigns and I also collaborate with the other PhD students at my lab. I enjoy the exchange and also the uplifting feedback, if some experiments go wrong, the small help for example if you (especially as a beginner) are not sure about some specific terminologies in IR or I enjoy to exchange about current tasks or general trends in IR and our research.

S: What excites you the most about the field of IR?

S: I am most excited about bringing advances from NLP to neural IR and furthermore transferring advances from web search to professional search by taking into account the special characteristics of professional search and the special domain. It excites me that the research area here is so vivid with so many new published ideas, also including so many different dimensions of IR like explainability, bias, computational complexity, speed or different aspects of relevance.

S: Walk us through a day in the life of a DoSSIER PhD student.

S: At the moment I am lucky that I can go to the office! So I usually get up at 8am, get dressed, no coffee, and then hop on my bike to drive to work (its only 5 minutes downhill). At university I have my own office which I share with another PhD. The first thing I do in the morning is that I check my emails and I check arxiv for new publications in IR. I try to work on one topic a day, so I either code or write in the morning. At roughly 12 me and my colleagues have lunch and coffee. Most of them are from Vienna, and that’s really cool, they tell me about local politics (mainly what they messed up) and what’s there to do in Vienna. After lunch I try to schedule most of my meetings or if I don’t have any, I continue with experiments or reading :)

This project has received funding from the European Union's Horizon 2020 research and innovation programme under the Marie Skłodowska-Curie grant agreement No 860721