Ben Thies

Ben Thies

Strategy Consultant / Researcher

Strategy&, Part of the PwC Network

MCC

Biography

Hi! I am Ben, a current Strategy Consultant at Strategy& and visiting scholar at the Mercator Research Institute on Global Commons and Climate Change’s Sustainable Resource Management and Global Change working group.

In the past, I leveraged my strong background in various social sciences paired with my formal training as statistician to research individual and collective human behaviour, and to develop computational research methods (as, e.g., here). I like to combine different methods and perspectives from other areas of research to my own work, I like to observe how humans interact, and I like to work with data.

This is what brought me to my current profession outside of academia. As a strategy consultant, I help clients approach their challenges from new angles. I specialize in topics revolving technology strategy (e.g. Data Governance, AI strategy, IT strategy), with a focus on the public sector. My hope is that with my projects I can make a positive contribution to a better, more digital, citizen-centric public administration.

I am always happy to collaborate on scientific research - or on discussions about the digitisation of the (mostly German) public sector - just hit me up!

Interests
  • Digitisation in the Public Sector
  • Behavioural Psychology
  • Data Science / Statistical Methods
Education
  • M.Sc. Statistics (GPA 1.0 | A+), 2021

    Humboldt University / Institute of Technology / Free University, Berlin | Duke University

  • B.A. Communication, Culture, & Management (GPA 1.1 | A+), 2018

    Zeppelin University, Friedrichshafen | Maastricht University

  • B.A. Corporate Management & Economics (Minor), 2018

    Zeppelin University, Friedrichshafen

Recent Publications

(2023). Machine Learning from Big GPS Data about the Heterogeneous Costs of Congestion. SSRN.

PDF Cite DOI

(2022). Using explainable machine learning to understand how urban form shapes sustainable mobility. Transp. Res. D.

PDF Cite DOI

(2021). Walking Through Twitter: Sampling a Language-Based Follow Network of Influential Twitter Accounts. SM+S.

PDF Cite Code DOI

(2020). The Importance of Suppressing Domain Style in Authorship Analysis.

PDF Cite

Talks

Mining Influencers in the German Twittersphere
This conference contribution presents a new sampling approach for large follow networks of influential Twitter accounts, discusses …
Mining Influencers in the German Twittersphere
Walking Through Twitter: How to Mine a National Follow Network
Twitter is used by individuals, grassroots movements, and political and social elites to directly communicate to the public and …
Walking Through Twitter: How to Mine a National Follow Network

GitHub & Coding

RADICES

The Rank Degree Influencer Core Sampler (RaDICeS) allows to draw an effective sample of a (language-based) Twittersphere as described in this talk. The underlying method is explained in this journal article. You can also cite the software itself.

Affinity Propagation Clustering

The affprop package is an Affinity Propagation Clustering (Frey and Dueck, 2007) implementation for Python. This package was part of a project of Cliburn Chan’s STA663 Statistical Computation class at Duke University.

trollR

trollR is an R package and tool to detect toxic language. It scored second place at the LSE Computational Social Sciences Hackathon ‘18.