event

UM Data Science Research Seminar Series

The Maastricht University’s UM Data Science Research Seminar Series brought together data scientists from across Maastricht University to discuss breakthroughs and research topics that are related to Data Science. We also welcomed interested students, colleagues and industry partners to experience bleeding edge presentations and discussion rounds. The aim was to increase awareness around the university about the data science research of its various departments and faculties, and to uncover new possibilities for collaboration.

With its inception in 2017, we gained experience in the do’s and don’ts of organizing these events. Following which, the 2018 edition of the Data Science Research Seminar Series was successful with 10 different departments participating across UM. We started in January 2018 by fixing the third Thursday of every month as the date for the session and sending out invites to our contacts in the different departments. Within almost a week, all the slots were filled by the departments, namely: M4I, BigCat, FaSoS (2 sessions), BISS, MAASTRO, GSBE, Faculty of Law, TGX and DKE.

The 10 sessions, which lasted 1 hour each, ranged in format from

  • Two talks by a student and a professor
  • A talk followed by a discussion on a hot topic
  • One 40 minute lecture followed by discussions
  • A debate, which was also streamed on facebook live
  • Talks by two assistant professors including Q&A

The topics of the sessions varied widely but all focused on the application of Data Science solutions to research problems. There were talks and discussions on topics ranging from Neural networks, WikiPathways, Crowdsourcing, Social sciences, Big data, Data integration, Data-driven decision making and Personalized medicine, Network analysis of court decisions, Data visualization, Machine learning and AI.

On average, 40 participants attended the sessions. Lunch was provided in each session and the cost was split between the Institute of Data Science and the collaborating department. As a result of the discussions, by recommendation by one of the attendees, a LinkedIn group was also formed so as to facilitate further discussion and exchange amongst participants beyond these sessions.

 We also received feedback from the participants who, in general, appreciated these events, had a chance to showcase their work “in progress” and get valuable insights for future directions relevant to their projects. Check out pictures and videos from all sessions here: https://www.facebook.com/pg/IDSatUM/photos/?tab=album&album_id=368703156871341

We will be continuing the series in 2019 with similar sessions with departments across UM and adding an expert lecture series as well. We advertise our events via our facebook, LinkedIn, twitter, instagram accounts as well as on the UM website. Please follow us to stay up-to-date of these events. Also, if you would like to be notified of upcoming events via email, please send an email to info dash ids at maastrichtuniversity.nl and we will sign you up in our mailing list. Looking forward to seeing you at the sessions this year !

wordle.png
Wordle from titles of all talks.
event, Uncategorized

 

Michel Dumontier, Alexander Malic, Vincent Emonet, Kody Moodley and Remzi Çelebi look back on this year’s SWAT4HCLS Conference (3 to 6 December in Antwerp, BE:
“This year’s conference put a spotlight on Agrisemantics. People shared their work and thoughts on various interesting topics including reproducibility, provenance, data integration, graph embeddings and privacy. The Institute of Data Science had the opportunity to present four talks*, a tutorial** and a hackathon proposal*** around a framework developed at the Institute. It enabled us to make new connections, learn new things and reconnect with old friends.”

* Talks:
1. A model for capturing provenance of assertions about chemical substances. Kody Moodley, Amrapali Zaveri, Chunlei Wu and Michel Dumontier. (https://bit.ly/2A0lt4U)
2. Evaluation of Knowledge Graph Embedding Approaches for Drug-Drug Interaction Prediction using Linked Open Data. Remzi Çelebi, Erkan Yaşar, Hüseyin Uyar, Özgür Gümüş, Oguz Dikenelli and Michel Dumontier.
(https://bit.ly/2A0lLJ2)
3. Vec2SPARQL: integrating SPARQL queries and knowledge graph embeddings. Maxat Kulmanov, Senay Kafkas, Andreas Karwath, Alexander Malic, Georgios V Gkoutos, Michel Dumontier and Robert Hoehndorf.
(https://bit.ly/2EywIVV)
4. Data2Services: enabling automated conversion of data to services. Vincent Emonet, Amrapali Zaveri, Alexander Malic, Andreea Grigoriu and Michel Dumontier.
(https://bit.ly/2ECzV6W)

** Tutorial:
Data2services, converting your data to a standard data model
presented by Vincent Emonet & Alexander Malic
(https://bit.ly/2BOabBU)
*** Hackathon challenge on Data2Services, presented by Vincent Emonet & Alexander Malic
(https://bit.ly/2Go1CC3)

#UMDataScience #DataScience #SWAT4HCLS

 

event

Workshop Ontologies – the building blocks of Tim Berners-Lee’s future vision for the World Wide Web!

On the 24 May 2018, the Institute of Data Science at Maastricht University (IDS@UM) held a workshop on Ontologies – the building blocks of Tim Berners-Lee’s future vision for the World Wide Web!

A truly diverse group of participants assembled in a remote, top-secret location (well, not really top-secret… it was Room G6.203 in the UNS50 building at Randwyck campus) to learn more about ontologies and their applications for solving problems in data-driven domains.  An intriguing mix of students and senior staff from different sides of the university (and beyond) were all united by their curiosity about the workshop topic. Datahub, MAASTRO, MUMC, UCM, ICTS and even KU Leuven were all represented.

We started with a gentle introduction to ontologies from the evolution of the term in philosophy to its use in information and computer sciences today. Then, the participants got “down and dirty” with actually building their own movie ontology with the Protege ontology editing software. After the lunch break we looked at several real-world applications of ontologies today. We discussed their extensive usage for standardising terminology in the healthcare and life sciences domains and also for representing terms in Google’s Knowledge Graph – more info hereFinally, we tinkered with an ontology-driven application for extracting insightful knowledge about human diseases. 

We wrapped up proceedings with a discussion on possible uses of ontologies in everyone’s research. Participants expressed appreciation, both during and after the workshop, for the clear presentation of the material and the novelty of the topic (some participants were introduced to ontologies for the very first time!) Overall, there was consensus that ontologies are useful for their future endeavours. Although, in some cases there were more concrete and immediate plans to apply them, and in others, more thought was needed to learn how to go about this.

This workshop was the latest edition in the Data Science Workshop Series organised by IDS. Workshop materials can be found on GithubPrevious editions of our workshop series dealt with topics ranging from Docker and Blockchain, to Machine Learning. 

The next workshop in the series will be provided by Dr. Amrapali J. Zaveri. on Crowdsourcing on June 12!

Hope to see you there!


event, think tank

Law/IDS/DKE Think Tank Meeting

The Faculty of Law (FL) and the Institute of Data Science (IDS) recently joined forces to hold a successful “Think Tank” meeting on the 17th of January. The event brought together colleagues from the Department of Data Science & Knowledge Engineering (DKE), FL and IDS to brainstorm about opportunities for educational and research collaboration.  

Graciously hosted and chaired by Marta Santos Silva (FL) at FL, enlivening opening addresses were also given by FL Dean – Jan Smits, and director of IDS – Prof. Michel Dumontier. Both talks ignited the enthusiasm of the Data Scientists and Law Professionals present to collaborate with each other in their respective projects.

25 participants from the three groups each presented a rapid-fire 2-minute introduction and pitch of their research to get acquainted with each other. The diversity in the room was plain to see, with attendance from Law and Computer Science Professors, Bachelors, Master and PhD Students, Postdocs, Software Developers and Data Scientists.

After the pitches, attendees were split into break out groups of 5 participants each and were set tasks to brainstorm on three topics of collaboration: education, research and infrastructure. The discussions were lively and fuelled by lots of coffee and delicious cake!

Proceedings were then rounded off with informal chats and networking over drinks and refreshments at FL. The organisers were overwhelmed with positive feedback from attendees, and are now arranging follow up discussions to further solidify the exciting cross-cutting research directions identified.

Watch this space for upcoming news about these discussions!

event

Data Science Symposium

On 25th January, 2018, we organized a Symposium on “The Future of a Data-Driven Society” at Maastricht University. With over 170 participants, the event was a huge success.

The symposium featured two keynote lectures by leading professors in the field of data science and digital technology:

  • Prof. Carole Goble from the University of Manchester set out her vision on the future of science. Prof. Carole Goble emphasised the importance of sharing and publicising research data according to the FAIR principles (Findable, Accessible, Interoperable, Reusable), and demonstrated which steps she takes to do so.
  • Prof. Lucy Suchman from Lancaster University. Prof. Lucy Suchman is an anthropologist with extensive experience and expertise relating to the interaction between people and computers. She explained her recent study into the role of robots in geriatric care, and presented a new approach to technological design within the field of healthcare.

There were two competitions and the following teams won:

  • Research competition for exceptional research proposals that focus on the use of data science to tackle problems of significant relevance to society. We received 13 submissions, 3 of which were invited to present at the symposium. The proposal entitled “Intelligent games for assessing cognitive, social, and physical capabilities of elderly and children” won this competition. You can see all the pitches here
  • Vision essay competition for visionary essays concerning the future of a data-driven society. We received 9 submissions and the winner was Claudia Egher with her essay titled: “Digital morning or manic by design.”. You can see all the submissions here.

Additionally, there was a poster session where we received 18 submissions from 11 different departments. The presenters had a chance to interact with the participants before and after the talks. This provided an opportunity to showcase innovative projects to the data science community.

We received overwhelming response from the participants (both present and remote) and requests to organize the event annually. A new website for UM’s data science community was launched to foster collaborations between different departments at Maastricht University. You may watch the video of the entire event here: https://www.youtube.com/watch?v=JN9eMMtCHf8.

Follow the hashtag #UMDataScience and Like us on facebook to keep updated on events and news and see you next year at the symposium !