Ruprecht-Karls-Universität Heidelberg

Chair of Computational Linguistics

Prof. Dr. Anette Frank

Anette Frank

Contact

Institut für Computerlinguistik        email: frank sign cl sign uni-heidelberg sign de
Im Neuenheimer Feld 325phone: +49-(0)6221/54-3247
D-69120 Heidelberg sek: +49-(0)6221/54-3245
Germany fax: +49-(0)6221/54-3242

Research

My research focuses on statistical - nowadays mostly neural - NLP methods for discourse-oriented semantic analysis and information access tasks with the aim of improving computational natural language understanding.

For more information please consult my research and publication pages and the pages of my research group.

I am member of the standing reviewer committee of TACL.

Office Hours and Teaching

Job Applications

I am always interested in highly qualified doctoral students, and I do look at the applications sent to me!
Please understand that I am not able to respond to individual inquiries, and do not send reminders.

News

  • Two long papers by members of the NLP group have been accepted for EMNLP 2020
    • X-SRL: A Parallel Cross-Lingual Semantic Role Labeling Dataset, by Angel Daza and Anette Frank, to appear in the Proceedings of EMNLP. The publication is based on work within the LiMo project
    • Social Commonsense Reasoning with Multi-Head Knowledge Attention, by Debjit Paul and Anette Frank, to appear in Findings at EMNLP.
  • The paper Argumentative Relation Classification with Background Knowledge, by Debjit Paul, Juri Opitz, Maria Becker, Jonathan Kobbe, Graeme Hirst, and Anette Frank earned a nomination for the best student paper award at the COMMA 2020 Conference! Congratulations to all authors!
  • Three new publications by members of the NLP group and colleagues:
    • Juri Opitz, Letitia Parcalabescu and Anette Frank (2020): AMR Similarity Metrics from Principles. Transactions of the Association for Computational Linguistics, Vol. 8, pp. 522-538. [html]
    • Debjit Paul, Juri Opitz, Maria Becker, Jonathan Kobbe, Graeme Hirst, and Anette Frank (2020): Argumentative Relation Classification with Background Knowledge. Proceedings of the 8th International Conference on Computational Models of Argument (COMMA 2020) [html]
    • Maria Becker, Ioana Hulpus, Debjit Paul, Juri Opitz, Jonathan Kobbe, Heiner Stuckenschmidt and Anette Frank (2020): Explaining Arguments with Background Knowledge. Towards Knowledge-based Argumentation Analysis. Datenbank Spektrum 20:131–141, Special Issue: Argumentation Intelligence. [DOI]
  • In the newest issue of Heidelberg University's research magazine on “Machine & Human”, Anette Frank and Katja Markert discuss "The Challenges of Teaching Machines to Understand Human Language". They discuss current work of their research groups and argue that both commonsense knowledge and creativity - which humans master so easily - are still a challenge for automatic language processing algorithms. Enjoy!

  • New publication: Letitia Parcalabescu and Anette Frank: Exploring Phrase Grounding without Training: Contextualisation and Extension to Text-Based Image Retrieval. Workshop on Multimodal Learning (at CVPR 2020), [pdf].
  • I will be Area Chair in the track Semantics: Sentence level, Textual Inference and Other area" for EMNLP 2020. Looking forward to a great conference!
  • I will be Area Chair in the track Textual Inference and Other Areas of Semantics for ACL 2020 in Seattle.
  • Letitia Parcalabescu and myself are participating in the newly established COST Action Multi3Generation: Multi-task, Multilingual, Multi-modal Language Generation, led by Isabelle Augenstein. We are members of the COST Action's management committee as representatives for Germany. Looking forward to collaborating in this great network activity.
  • Three members of the NLP group will attend EurNLP 2019 in London to present and discuss their research:

    • 
Todor Mihaylov: Discourse-Aware Semantic Self-Attention for Narrative Reading Comprehension
    • Letitia Parcalabescu: Unsupervised Text to Scene Graph Alignment with WordNet Neighborhood
    • Debjit Paul: Graph-based Multi-Hop Commonsense Knowledge

Selected new publications

Current Projects

Completed Projects

Former Affiliations

zum Seitenanfang