| layout | page |
|---|---|
| title | Services |
| menu | main |
| navigation_weight | 3 |
| color | #f4cae4 |
Design session at CODATA-RDA Summer School, Trieste (Italy), Summer 2016
We are available for lectures/workshops and development of:
- custom visual design for complex data
- prototype development for exploration of complex data
- data visualisation as a tool for thought
- topological data analysis of domain-specific data
Custom visualisations in life sciences
Topological network of air quality in northern Spain
{% comment %}
The UHasselt Data Science Institute in Belgium (www.uhasselt.be/dsi) fosters collaboration among data science experts across disciplines. As part of this new institute, the Visual Data Analysis Lab is interested in applying interactive data visualisation to support data scientists in gaining insight in their complex data.
To further extend our expertise in contextualisation of data analysis, the lab will be hiring a doctoral student in the field of topological data analysis where we convert high-dimensional datasets into networks in order to extract meaningful patterns. More specifically, this involves research in the definition of feature vectors in multiple domains, graph distance metrics, and generation and qualitative analysis and comparison of networks. Application domains include life sciences (epidemiology, genomics, patient stratification, ...), archaeology and others.
- A strong MSc degree in computer science or design, or a field related to human-computer interaction
- Some experience in visual design or data visualisation (e.g. using D3)
- Good programming skills (preferably python and/or javascript) and algorithmic thinking
- Fluent in spoken and written English
The successful candidate will be appointed for a period of 2+2 years at UHasselt.
For more information regarding this function and to apply, please go to https://www.uhasselt.be/vacancies_detail?taal=04&vacid=1836.
Recent advances in deep learning are leading to applications ranging from automatic text translation and machine reading, to style transfer (e.g. between paintings, but also applied in medical imaging) and self-driving cars. A significant drawback of deep learning networks, however, is that they act as black boxes with very limited transparency and therefore the user needs to have blind trust in whatever answer is returned to them.
The Visual Data Analysis Lab - embedded in the UHasselt Data Science Institute in Belgium (www.uhasselt.be/dsi) - is interested in the application of interactive data visualisation to support researchers in the pursuit of deep insights in complex data. We employ topological data analysis (see our work in review at https://arxiv.org/abs/1907.05783) and interactive data visualisation to identify hidden patterns in highly dimensional data in domains ranging from life sciences to archaeology and bibliometrics, and aim to use this approach towards advancing the field of explainable AI. Therefore, we seek to employ an enthusiastic doctoral researcher, who will focus on interactive data visualisation and development of new algorithms for creating and comparing the topological graphs in the context of deep learning, as well as perform data analysis, write research papers, and present his/her work at national and international conferences.
- You have obtained a Master in Informatics or Visual Analytics (or equivalent)
- You have a clear affinity with visual analytics and/or data visualisation, ideally with expertise in D3.
- You are familiar with techniques for deep learning, data clustering and dimensionality reduction
- You have good programming skills (preferably python and/or javascript), including scalability of algorithms
- You are good in algorithmic thinking.
- You are able to work in collaboration with domain experts as well as work independenlty
- You are good in clear and regular reporting.
- You have high proficiency in English, both written and oral.
The successful candidate will be appointed for a period of 2+2 years at UHasselt.
For more information regarding this function and to apply, please go to https://www.uhasselt.be/vacancies_detail?taal=04&vacid=1864.
Although different generic data visualisation tools exist (such as Tableau and Spotfile), there is a specific need for custom visualisation (eg hive plots, horizon charts, and plots that incorporate domain-specific information) when working with complex datasets. In this project we will build a framework for creating custom visualisations and combining these in a dashboard-setting. The aim is to create the necessary front- and backend to allow for (1) easily creating and incorporating new visual modules (conceptually similar to https://github.com/jandot/bio-graphics), (2) combining these visual modules in a single interface, and (3) incorporating a biological ontology which links biological concepts to features in the visuals.
The code developed in this project will be partly open sourced, and we aim for reaching a large community in the biological sciences initially, but later also in other domains.
The candidate will work in close collaboration with a genomics researcher, and be responsible for:
- Analysis and deconstruction of prior work
- Co-develop data schema, develop data storage solution, and data loading scripts for omics-specific data
- Develop proof-of-concept interface which links different visualisations
- Implementation of custom visuals designed by visualisation experts
We offer a challenging position within a dynamic and stimulating environment and ask for a highly motivated, enthusiastic and pro-active person. The applicant needs to have demonstrable experience in web-development and software management. This includes strong programming skills in javascript and if possible python, and/or clojurescript.
The successful candidate will be appointed for a period of 2 years at KU Leuven, starting at or as soon as possible after April 1st 2019. Salary will depend on diploma ("barema") and experience ("trede"). Barema 44 applies to applicants who completed a doctorate; barema 43 applies to applicants without.
For more information regarding this function, please contact jan.aerts@uhasselt.be or vera.vannoort@kuleuven.be.
Please send - in PDF format:
- curriculum vitae
- motivation letter
- a brief description how you would approach this problem, and which technologies you would use
- link(s) to work that you already did regarding web-development
The Visual Data Analysis Lab and Research[x]Design Lab at the KU Leuven (Belgium) have a joint open PhD position in the field of data science and visualization / interaction design. The candidate will be part of a multidisciplinary team of researchers around the Sagalassos Archaeological Research Project, covering the domains of archaeology, geography, ecology and spatial planning.
We seek to unravel the interdisciplinary understanding of change and its space/time dimensions within already existing and future data sources of the Sagalassos project (running since 1986), which investigates an archaeological site in Turkey and studies its uninterrupted occupation over nearly 2,000 years in all its aspects, from daily life, diet, religious beliefs, to architecture, production, trade and its mechanisms, as well as natural resources and environmental conditions of the past.
This is a joint project between research groups of archaeology, ecology, geography, architecture and data analysis/visualization (STADIUS and RxD). The candidate will be supervised by Profs Aerts and Vande Moere at STADIUS and RxD, respectively, in close collaborations with the other partners in the project.
Links:
- Research[x]Design: http://rxd.architectuur.kuleuven.be/
- Visual Data Analysis Lab: http://vda-lab.be
- Sagalassos Archaeological Research Project: http://www.sagalassos.be
Candidates must have a Master's degree or equivalent in computer science, computational design or a related field, and have very good knowledge of data visualization or data analytics, and human-computer interaction, or demonstrate the ability to acquire sufficient skills in all of these areas. Solid programming skills and fluency in spoken and written English are required and could be evidenced by examples of working visualisations.
Tasks will include the user-centered and iterative design of visualisations that are able to meaningfully combine inherently disparate data sources from different disciplines (i.e. archaeology; geography; ecology; spatial planning) that are involved in the Sagalassos project, and the evaluation of these visualisations in terms of (synchronous) cross-disciplinary and collaborative insight and hypothesis generation.
As such, the candidate is expected to merge the expertise from the fields of human-computer interaction (i.e. user studies, interaction design, prototyping, evaluation) with that of visual analytics (i.e. combining representation with machine learning; incl. clustering and dimensionality reduction). Depending on the potential interaction requirements, explorations towards promising emerging technologies such as augmented reality (AR) and/or physicalisation are possible.
Please note short-listed candidates might be asked to fulfil concise assignments to test their proficiency.
We offer:
- Full-time employment starting January 1st 2018 for a period of four years, with the aim of obtaining a PhD degree
- A competitive salary
- A stimulating scientific working environment, within one of Europe's leading research universities
For more information please contact Prof. dr. ir. Jan Aerts, tel.: +32 16 32 10 53, jan.aerts@kuleuven.be or Prof. dr. Andrew Vande Moere, tel.: +32 16 32 13 62, andrew.vandemoere@kuleuven.be. Please apply through the KU Leuven open positions website
Patient records and clinical reports include a vast amount of crucial information that is stored as free text, but that is unstructured and requires a human to make sense of it (e.g. “vague pain in belly which radiates to back”). Research into semantic processing of these texts (“machine reading“) is a hot topic at the moment and aims to not merely extract the individual components of these texts, but actually come to a deeper machine “understanding”, including temporal, spatial and causal structures and relationships.
The ACCUMULATE consortium recently received 4-year funding for “Acquiring Crucial Medical Information Using Language Technology”. Our role in this project is to develop data visualization methods to (a) support the development of the machine reading algorithms, and (b) present the extracted information to the doctor. For the first, we will for example create abstract interactive visualizations showing the effects of using different parameters in the algorithms; the latter will include (but not be limited to) creating visuals that combine the information from the unstructured text with additional information such as episodes, medication, etc. To do this, we have 2 fully-funded positions available at the moment, for a 4-year PhD student and a 1-year postdoctoral researcher. Both positions will include working with the users, visual design, implementation of data visualizations, and evaluation.
- A strong MSc degree in computer science or design, or a field related to human-computer interaction
- Some experience in visual design or data visualization (e.g. using D3), preferably of text
- Good programming skills (preferably javascript)
- PhD which combines computer science with visual design
- Demonstrably track record in visualization of textual information
To apply, please send an email to jan.aerts@kuleuven.be, and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] one or more figures and/or screencasts of previous data visualization work. We will start reviewing applications in October 2015, but applications will be considered until the positions are filled.
Did you ever want to use your design background to save lives? We're hiring!
Cancer- and epilepsy-patients as well as early-born babies are a few examples of people who are in the care of entire teams of specialists, ranging from radiologists and oncologists to cardiologists, neurologists and other -ologists. Communication between these experts is of course crucial, but the tools that they use for this often date from the stone age (powerpoint with screenshots, for example.)
We want to change that.
In collaboration with companies such as Barco, which develops big screen hardware, we want to focus on the design/visualization-end of these interactions. Some questions that we would like to answer include: To what extent can data from one field be abstracted so that an expert in another field can still understand it? What would a good interactive expert-oriented visual overview look like that can be used as a scaffold to attach more detailed information? Can we find a way to gracefully degrade that overview in settings where one of the collaborators works remotely on a small or low-resolution screen? Can we integrate for example neurology data with cardiology?
But we need your help...
If you have a design background and know your way around user-driven design and programming data visualizations using Processing or D3, you may be the person we are looking for. If you're interested to play a role, please send an email to jan.aerts@kuleuven.be, and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] one or more figures and/or screencasts of previous data visualization work. Please mention "MECOVI" in your application. We'd like to find someone by February...
In case you have an MSc, the position is initially for 15 months, but if we're a good match additional funding will be sought to prolong this for a full PhD study. Postdoctoral researchers are free to apply as well, but in that case funding will be for a shorter period.
Keywords: visualization, collaboration
In the old days, all information relevant to a patient would have been hand-written on pieces of paper stored in the doctor's office. Things have changed. Not only is there the electronic patient record containing an individual's full history, but a lot of other relevant information is available such as medical guidelines, aggregated characteristics of comparable patients, and projections into the future. Wouldn't it be nice if we could let patients/citizens give full access to and insight in their own data? We believe so...
In collaboration with different Belgian content providers (electronic health record providers, caregivers, health insurers, ...) we want to build tools to unlock this information to the patient. This will involve data analytics, visualization, and user experience research. To help us in this effort, we are looking to hire a data scientist versed in the field of data visualization but also not shy to perform data analysis. Specific topics that will be covered include (but will not be limited to) data exploration, improving interpretability of black-box machine learning algorithms, and development of privacy-aware visual encodings.
So if you're a data scientist who not only knows how to cluster datasets but also how to design visualizations (e.g. using Processing or D3, you might be the person we're looking for. If you're interested to play a role in this project, please send an email to jan.aerts@kuleuven.be, and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] if available, one or more figures and/or screencasts of previous data visualization work. Please mention "MyHealthData" in your application. We'd like to find someone by February...
If you have an MSc, the position is originally available for 2 years, but in case of a good match additional funding will be sought for PhD study. If you're a postdoctoral researcher, you're welcome to apply as well, although the funded period will initially be shorter.
Keywords: visualization, healthcare
Viruses are the cause of many common human diseases, from the not-so-serious as common cold and cold sores, to the very-serious like HIV and Ebola. Unfortunately, the tools to identify and follow the spread of these viruses leave a lot to be improved. In the European Virogenesis project, we want to change this by building the bioinformatics tools to diagnose viral diseases, monitor their evolution within patients, and track viral outbreaks (think "Ebola").
Therefore, we are looking to hire a PhD student to focus on the visualization aspects of this project. This includes (but will not be limited to) addressing scalability, uncertainty, and interactivity of phylogenetic trees. In addition, we need to devise ways to combine these trees with for example geo-location data, and find visual encodings to represent changes in viromes.
Initial funding is available for 3 years. Additional funding will be sought to extend this to 4 full years. Postdoctoral researchers are welcome to apply as well, but initial funding will be for a shorter period.
Are you interested in helping us out and have a background in omics data analysis and/or data visualization (e.g. using Processing or D3? If so, please send an email to jan.aerts@kuleuven.be, and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] if available, one or more figures and/or screencasts of previous data visualization work. Please mention "Virogenesis" in your application.
Keywords: metagenomics, phylogenetics, visualization, bioinformatics
Imagine being an air traffic controller and wanting to know if a plane is not following the route it should be following. Or imagine being a maritime law enforcement officer and wanting to identify which ships in the Indian ocean might be pirate ships. This is surprisingly difficult to do, because you can't always know beforehand what to look for. This is where geo-spatial data visualization comes in, showing the trajectories of these planes or ships on a map in combination with additional data.
But is it that simple? No. How can you find that one ship that you're interested in amongst the thousands of ships on your screen? In this project, we would like to develop visual analytics techniques for easy identification of "abnormal" behaviour in geo-spatial data, in collaboration with Luciad, a Leuven-based company specialized in situational awareness that helps their users to gain insight in large time-dependent geo-spatial data.
Ship movements at port of Rotterdam. van de Laar et al, Situation Awareness with Systems of Systems
Therefore, Luciad and our lab are looking to hire a PhD researcher under the conditions of a IWT Baekeland grant, which is a personal scholarship assigned to you. The research would be in very close collaboration with Luciad.
So if you're a data scientist who has some background in machine learning (clustering, ...) and is interested in learning to create interactive data visualizations (e.g. using Processing or D3, you might be the person we're looking for. If you're interested to play a role in this project, please send an email to jan.aerts@kuleuven.be (cc to bart.adams@luciad.com), and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] if available, one or more figures and/or screencasts of previous data visualization work. Please mention "GeoPattern" in your application. Deadline for the grant proposal submission is beginning of March, which means we need someone in the first half of February. The grant procedure includes an oral defense in Brussels by the candidate, so it helps if you're from Belgium or not that far away.
For years, we thought that single small changes in the DNA (single nucleotide polymorphisms) were the main actors in what distinguished the diseased from the healthy, the susceptible from the resistant. But now it becomes more and more clear that larger structural genomic variations (insertions, deletions, inversions and translocations) play an important role as well. Why were these hidden for so long? Because the data is more difficult to find and analyze. Nevertheless, their analysis comes more and more under pressure because people start looking at the single-cell level, for example to see how tumours evolve over time.
To help the analysis and interpretation of this type of data, we need to work on methods for visually representing these variations. And for that, we are looking to hire a PhD student under the Brazilian Science Without Borders (Ciência sem fronteiras) umbrella.
Funding for this project is provided by the Brazilian government, which means that only people with the Brazilian nationality can apply.
Do you have the Brazilian nationality? And do you have a background in data visualization (e.g. using Processing or D3), or know your way in genome analysis? Then you might be the person we're looking for. If you're interested, please send an email to jan.aerts@kuleuven.be, and include (preferably in PDF): [a] your CV, [b] two references (with email and phone number), [c] if available, one or more figures and/or screencasts of previous data visualization work. Please mention "ScienceWithoutBorders" in your application.
Keywords: genomics, visualization, structural variation {% endcomment %}