Andreas Stephan

Andreas Stephan

PhD student in Natural Language Processing

University of Vienna

Biography

Hey! My name is Andy, and I am currently pursuing a PhD with the Digital Text Sciences group under the leadership of Professor Benjamin Roth at the University of Vienna. The group is part of the larger Research Group Data Mining and Machine Learning. My research focuses on leveraging various noisy or weak signals to enhance or direct learning algorithms. This includes working with labeling functions—code that annotates data, image-to-text models providing imperfect descriptions of images, and outputs from multiple large language models (LLMs). Prior to my academic engagement, I spent two years in the industry, tackling applied natural language processing (NLP) challenges, including information extraction and the integration of graphs with textual data.

Download my resumé .

Interests
  • NLP (in general)
  • Weak Supervision
  • Multi-modality
  • Multi-source information
  • Mutli-Agent
Education
  • PhD in NLP, 2021 -

    University of Vienna

  • M.Sc. in Mathematics in Data Science, 2019

    Technical University Munich

  • B.Sc. in Mathematics, 2017

    Technical University Munich

  • B.Sc. in Computer Science, 2014

    Technical University Munich

News

Two oral presentations at EACL 2024

Short paper accepted at "The Web Conference"

On month research stay at the Schütze lab at LMU in Munich, Germany.

Invited talk at Munich NLP

Paper accepted at EMNLP 2023

Weak Supervision Tutorial at Aalborg University Copenhagen, Denmark

Our Paper obtained the ACL 2023 Theme Paper Special Award

Invited talk at Snorkel ML

I gave a talk at VW Data labs, Munich

Selected Publications

(2024). Text-Guided Image Clustering. In EACL-2024.

PDF Cite Code ArXiv

(2024). Counterfactual Reasoning with Knowledge Graph Embeddings. In EACL-2024.

PDF Cite Code ArXiv

(2023). Seeing through the mess: evolutionary dynamics of lexical polysemy. In EMNLP-2023.

PDF Cite

(2023). Towards Enhancing Deep Active Learning with Weak Supervision and Constrained Clustering. In IAL@ECML-PKDD-23.

PDF Cite

(2023). Weaker Than You Think: A Critical Look at Weakly Supervised Learning. In ACL-2023.

PDF Cite

Teaching

YearCourses
SS 24Practical Machine Learning for NLP
Modelling and Handling of Large Databases
WS 23/24Deep Learning for Natural Language Processing
SS 23Scientific Data Management
WS 22/23Deep Learning for Natural Language Processing
SS 22Introduction to Mathematics for Computer Scientists
WS 21/22Seminar: Weakly Supervised Learning