Datenbanken und Informationssysteme

Ein Interview mit Eleni Tzirita Zacharatou (Englisch)

"DIMA is a very large data systems group, much larger than most others. As a result, it offers a stimulating environment with plenty of opportunities for growth through the interaction with fellow group members and undertaking different roles within the group. Furthermore, DIMA focuses on building full-fledged data processing systems that address real-world demands, which is challenging to achieve in smaller groups. Finally, the group has a strong presence in the international database community, publishing in top-tier venues every year."

1. How did you learn about DIMA? What was your first impression?

I first remember encountering a group of DIMA researchers at VLDB 2018 in Rio de Janeiro, Brazil when they visited my poster presentation. Our discussion was very interesting, and I was intrigued by their insightful questions. As I approached graduation and began searching for postdoc opportunities, several colleagues recommended DIMA to me, which confirmed my initial positive impression of the group. 

2. What were you working on during your Postdoctoral Research at DIMA?

During my roughly three years at DIMA, I had the opportunity to collaborate with many students and colleagues on various exciting research problems that were mostly orthogonal to my doctoral research.

I was part of the AgoraEO project, which started as a joint effort between the DIMA and RSiM groups at TU Berlin. To that end, I contributed to the design of AgoraEO, a platform that enables the discovery, sharing, and analysis of Earth Observation (EO) data [1]. Additionally, I worked on developing a search engine for AgoraEO that supports content-based retrieval over satellite image repositories [2].

I was also involved in mentoring doctoral students. Specifically, I co-mentored Xenofon Chatziliadis on the monitoring of stream processing engines in the context of NebulaStream [3,4]. Additionally, I collaborated with Martin Kiefer, a last-year DIMA doctoral student at the time, on a VLDB paper proposing an optimistic data parallel architecture for sketching on FPGAs, which allows for efficient resource utilization while mitigating resource conflicts [5]. 

I also supervised several Master’s and Bachelor’s theses on various data management topics. Notably, I collaborated with Felix Henze and Haralampos Gavriilidis on designing a spreadsheet parser that enables high-performance spreadsheet loading in data science environments [6]. Furthermore, I worked with Elena Beatriz Ouro Paz on a fault tolerance strategy that makes it possible to run stream processing analytics on mobile devices [7]. Lastly, I co-mentored Marcellus Prama Saputra on devising an in-place update algorithm for compressed bitmap indexes [8].

Finally, I was able to crystalize my vision, which was shaped during my Ph.D., of incorporating distance-bounded spatial approximations in the design of approximate spatial data processing techniques [9].

3. How would you differentiate DIMA from other research groups in Germany or across Europe?

DIMA is a very large data systems group, much larger than most others. As a result, it offers a stimulating environment with plenty of opportunities for growth through the interaction with fellow group members and undertaking different roles within the group. Furthermore, DIMA focuses on building full-fledged data processing systems that address real-world demands, which is challenging to achieve in smaller groups. Finally, the group has a strong presence in the international database community, publishing in top-tier venues every year.

4. How did DIMA help you to advance your career?

During my time at DIMA, I expanded my research skills, acquired new technical knowledge, gained experience in both the supervision and teaching, and strengthened my project and time management skills. This multi-faceted growth gave me the confidence to take on my next (i.e., current) role as an Assistant Professor at the IT University of Copenhagen.

5. What advice would you give to future Postdocs who want to enter the job market?

Here are a few suggestions based on my own personal experience:

  • A postdoc offers a unique chance to challenge yourself and operate outside of your comfort zone. In contrast to doctoral studies, as a postdoc you have more flexibility to explore different projects and collaborate more broadly with other researchers. You also have more time to focus on research than a junior faculty member would and no pressure to secure funding.
     
  • Consider doing your postdoc in a group that is different from where you completed your Ph.D., in terms of both the research focus and management style. This will give you a diverse perspective that will help you shape your own research group in the future.
     
  • Engage in activities that will help you prepare for an academic position, such as teaching, mentoring, reviewing, and organizing professional events. However, producing high-quality research and strengthening your publication record should be the priority.
     
  • Be intentional about celebrating your successes, as this can help you stay motivated and resilient in the face of setbacks.
     
  • Learn to manage your time and resources effectively, so that you can take advantage of opportunities that interest you and align with your career goals while setting boundaries for things that don't bring value.
     
  • Networking is crucial, especially if you belong to an underrepresented group, and requires a conscious effort.
     
  • If your goal is to apply for a professorship, start preparing your application materials (mainly the research statement) well in advance. This will give you enough time to refine your application and be ready to apply when opportunities arise. Also, have a clear timeline in mind for the duration of your postdoc and the targeted application period.
     
  • If you are doing your postdoc in a new city or country, take the opportunity to immerse yourself in the culture and meet new people, as this will significantly contribute to your personal growth.

References

[1] “Agora-EO: A Unified Ecosystem for Earth Observation -  A Vision For Boosting EO Data Literacy -”. Arne de Wall, Björn Deiseroth, Eleni Tzirita Zacharatou, Jorge-Arnulfo Quiané-Ruiz, Begüm Demir, and Volker Markl, International Conference on Big Data from Space, Virtual Event, May 2021.

[2] “Satellite Image Search in AgoraEO”. Ahmet Kerem Aksoy, Pavel Dushev, Eleni Tzirita Zacharatou, Holmer Hemsen, Marcela Charfuelan, Jorge-Arnulfo Quiané-Ruiz, Begüm Demir, and Volker Markl,  Proceedings of the VLDB Endowment 15(12): 3646-3649 (2022).

[3] “NebulaStream: Complex Analytics Beyond the Cloud”. Steffen Zeuch, Eleni Tzirita Zacharatou, Shuhao Zhang, Xenofon Chatziliadis, Ankit Chaudhary, Bonaventura Del Monte, Dimitrios Giouroukis, Philipp M. Grulich, Ariane Ziehn, and Volker Markl, Open J. Internet Things 6(1): 66-81 (2020).

[4] “Monitoring of Stream Processing Engines Beyond the Cloud: An Overview”. Xenofon Chatziliadis, Eleni Tzirita Zacharatou, Steffen Zeuch, and Volker Markl, Open J. Internet Things 7(1): 71-82 (2021).

[5] “Optimistic Data Parallelism for FPGA-Accelerated Sketching”. Martin Kiefer, Ilias Poulakis, Eleni Tzirita Zacharatou, and Volker Markl, Proceedings of the VLDB Endowment 16(5) (2023).

[6] “SheetReader: Efficient Specialized Spreadsheet Parsing”. Haralampos Gavriilidis, Felix Henze, Eleni Tzirita Zacharatou, and Volker Markl, Information Systems, 2023.

[7] “Towards Resilient Data Management for the Internet of Moving Things”. Elena Beatriz Ouro Paz, Eleni Tzirita Zacharatou, and Volker Markl, Datenbanksysteme für Business, Technologie und Web (BTW), Dresden, Germany, September 2021.

[8] “In-Place Updates in Tree-Encoded Bitmaps”. Marcellus Prama Saputra, Eleni Tzirita Zacharatou, Serafeim Papadias, and Volker Markl, International Conference on Scientific and Statistical Database Management, Copenhagen, Denmark, July 2022.

[9] “The Case for Distance-Bounded Spatial Approximations”. Eleni Tzirita Zacharatou, Andreas Kipf, Ibrahim Sabek, Varun Pandey, Harish Doraiswamy, and Volker Markl, International Conference on Innovative Data Systems Research (CIDR), Virtual Event, January 2021.