Track B: Data and Network Science with Applications

Whether you'll work in business, public administration, health care, or scientific research, you will have access to more data and intelligent networked systems than any employee who came before you: most of the digital data points or network links (which can be posts on social networks, friendship links on the same social networks, GPS locations, digital newspaper articles, bank transactions, or URLs on the Web) were created in the last few years [IBM]. To obtain concrete value from these networks and data, for a business or society, you need to learn and (particularly) practice working with the latest tools, frameworks, models, and algorithms which deal well with data that is very large, comes in complex data structures such as graphs, is noisy, incomplete, untrustworthy, or doesn't have an underlying model. Importantly, we'll ideally come up with that concrete research question which needs you to mix some theory and some practice in the research process, and solves a problem with some impact, from any domain of your choosing.
Project topics
Here is a colourful list of ideas, targeting different algorithmic domains, and different application domains. You could build on these, or come up with your own proposals.
(1) Study a Twitter community of your choice, for example, the Dutch users who are politically active on Twitter. Calculate and interpret features of this network; is it small-world? Does the volume of political tweets accurately reflect political polls? Can Twitter activity predict election results? (domains: text mining or network science).

(2) Study the way social network users experience a polarized exposure to news posts; when users have a political bias, how does this bias influence how far (for example) fake news will spread? Model the propagation of biased content and contrast with existing models of influence propagation (domains: network science.
(3) Design and build software tools which assemble, data mine, or analyze genomic data (DNA) [Genomics, SeqAs] (domains: string mining or artificial intelligence, genomics).
(4) Study the traffic patterns or vehicles in a given location: a country or a city [Nationale Databank Wegverkeersgegevens,NYC-bikes,SF-bikes,NYC-Uber]. Answer questions such as: Can you predict future traffic? Can you predict traffic jams? Does traffic correlate with other variables (be they economical, social, or cultural) ? (domains: logistics, data mining).
(5) Improve on current techniques for sentiment analysis in email, newspaper, and tweet text [Hillary's emails] to determine and verify the timeline of social or political events, and thus build an 'automated history' for a city, country, or continent (domains: text analysis).
(6) Pick a network protocol for communicating messages between users; this can be a wired or wireless message-routing protocol, or a social network. Design methods to uncover its vulnerabilities: can one or more malicious users disrupt the communication of legitimate messages by employing attack techniques such as flooding the network? (domains: algorithms or artificial intelligence, networking).
(7) Pick any data-rich open problem which needs a machine learning algorithm, or any computationally hard problem which needs an approximation or optimization algorithm. Example of the latter: evolve (optimize) a robot design [Evolved soft robots] (domains: machine learning, artificial intelligence).


[IBM] "90% of the data in the world today has been created in the last two years alone. This data comes from everywhere: sensors used to gather climate information, posts to social media sites, digital pictures and videos, purchase transaction records, and cell phone GPS signals",

[Nationale Databank Wegverkeersgegevens]

Further Information
For further information on the content of this track, you may contact the track chair Doina Bucur, For any information on the conference organisation, please contact the conference chair on