I am the CAROLL Research Group Leader (supported by the BMBF program for Women in AI), at the Faculty of Computer Science and Mathematics of the University of Passau, Germany.

In April 2023, I took a role of interim Professor of Legal Informatics and NLP at the Faculty of Law, University of Passau.

My research interests lie in the broad area of Natural Language Processing (NLP). I am especially interested in Computational Rhetoric, i.e. processing of non-literal language and detection of rhetorical figures, applied to Sentiment Analysis, Offensive language modelling and detection, Legal Tech, Argumentation Mining and other areas of NLP.

I live in Passau with my husband, two children, the world's calmest Jack Russell Terrier, Penny, and her younger, not so calm protégé, Bella.

My name is Serbian and has a diacritic at the last letter: Jelena Mitrović (Mitrovi\'{c})

• For all recent updates, please refer to the CAROLL Research Group page.
• The State-of-the-Art neural language model for abusive language detection in English HateBERT is now available via the HuggingFace Transformers Library (01.02.2022).
• In our new KI-Campus Research Project at the University of Passau called Deep Write we will build AI tools to help students of Law and Economics using argument mining methods (15.12.2021).
• The project SMAEG Bot für regionale Lebensmittel funded by the Federal Ministry of Food and Agriculture BMEL has officially started. The project aims to make the market for regional Bavarian food products more transparent and to reduce the marketing effort for the involved parties.
• I gave a Keynote Speech at The French-German Summerschool on AI with Industry 2021 organized by the University of Passau, Ecole Normale Supérieure Paris-Saclay and with the support from Siemens AI Lab.
• My BMBF project CAROLL was covered in Research in Bavaria
• The 17th edition of the MWE workshop will be collocated with ACL 2021
• I have been appointed to the Board of Scientific Directors of the Rhetoricon project (University of Waterloo, Canada) to help move the field of Computational Rhetoric forward.
• From October 1st 2020 I will be taking on the role of an interim Junior Professor of Computational Rhetoric and NLP, at the University of Passau.
• We've had 9 successful system submissions at OffensEval 2020 as part of SemEval 2020, some of which were ranked in the top 10 percent (system description papers pending publishing).
• Our team GruPaTo (with Uni Groningen and Uni Turin) at SemEval-2020 system submission paper is available as a preprint.
• Our team NLP_Passau at SemEval-2020 system submission paper is available as a preprint.
• Invited speaker at the Waterloo.ai Seminar, University of Waterloo, Canada, March 2020.
• Invited speaker at the Passau Data Science Summit, PasDas, organized by the University of Passau and ONE LOGIC, April 2020.
• I am a PI in the 2 year DAAD funded programme "Cross-Lingual Hate-Speech Detection" between the University of Passau and University of Belgrade.
• I am co-organizing a Joint Workshop on Multiword Expressions and Electronic Lexicons (MWE-LEX 2020)at COLING 2020 in Barcelona
• I co-organized a Joint Workshop on Multiword Espressions and WordNets (MWE-WN 2019)at ACL 2019
MWE-WN Proceedings
• I am a Management Committe member in the new COST action NexusLinguarumCA18209 - European network for Web-centred linguistic data science
• Our Submission system at SemEval 2019 task 6, OffensEval was ranked in the top 10 percent.

• We participated in the PAN CLEF Challenge on Bot Detection and Gender Profiling, ranking 10th (PAN CLEF 2019 Proceedings

• Analysis of a German Citation Network was published in Proceedings of the 13th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management - KDIR
• A Stacking Approach for Cross-Domain Argument Identification was published in The Proceedings of the 32nd International Conference on Database and Expert Systems Applications DEXA 2021.
• DistilBERT-based Argumentation Retrieval for Answering Comparative Questions was published in The Proceedings of Conference and Labs of the Evaluation Forum CLEF 2021.
• Ten Assumptions About Artificial Intelligence That Can Mislead Patent Law Analysis in collaboration with the Max Planck Institute for Innovation and Competition was published in SSRN Series.
• Our paper HateBERT: Retraining BERT for Abusive Language Detection in English was accepted for publication at the 5th Workshop on Online Abuse and Harms (WOAH) ACL-IJCNPL 2021.
• Our Paper Automated identification of bias inducing words in news articles using linguistic and context-oriented features was published in Elsevier's Journal of Information Processing & Management
• Our Paper Multi-word Expressions for Abusive Language Detection in Serbian was accepted for publication at the MWE-LEX Workshop, collocated with COLING
• Our paper Towards Classifying Parts of German Legal Writing Styles in German Legal Judgments was accepted for publication at the ACIT 2020 IEEE conference.
• Our paper I Feel Offended, Don’t Be Abusive! Implicit/Explicit Messages in Offensive and Abusive Language was accepted at the Main track of the LREC 2020 conference. The corresponding dataset can be found in the GitHub repository.
• Our paper was accepted for publication at the LREC 2020 conference in the the first edition of the REPROLANG track, Language Proficiency Scoring, Cristina Arhiliuc, Jelena Mitrović, Michael Granitzer
• Our paper was accepted for publication at the ICAART 2020 conference Cognitive Modeling in Computational Rhetoric: Litotes, Containment and the Unexcluded Middle, Jelena Mitrović, Cliff O'Reilly, Randy Allen Harris, Michael Granitzer
• nlpUP at SemEval-2019 Task 6: A Deep Neural Language Model for Offensive Language Detection, Jelena Mitrović, Bastian Birkeneder, Michael Granitzer
• upInf - Offensive Language Detection in German Tweets, Bastian Birkeneder, Jelena Mitrović, Julia Niemier, Leon Taubert, Siegfried Handschuh
• Bots and Gender Profiling using a Multi-layer Architecture, Notebook for PAN at CLEF 2019 Régis Goubin, Dorian Lefeuvre, Alaa Alhamzeh, Jelena Mitrović, Elod Egyed-Zsigmond, Leopold Ghemmogne Fossi
• Ontological representations of rhetorical figures for argument mining, Jelena Mitrović, Cliff O’Reilly, Miljana Mladenović and Siegfried Handschuh
• A cross-linguistics study on Greek and Serbian fixed similes and enrichment of lexical resources via crowdsourcing, Jelena Mitrović, Stella Markantonatou, Cvetana Krstev

LinkedIn


Google Scholar

Research Gate

CAROLL Research Group

• BMBF KI Campus funding scheme - PI in the University of Passau project Deep Write - duration 4 years, project start December 2021.
• BMBF funded Junior Research Group Leader - CARROL group, Computational Rhetoric in Social Media and Law, 1,9 M euro in funding, 4 years, project start October 2020 .
• PI in the DAAD Exchange Program between Serbia and Germany for Multilingual Hate-Speech Detection, HasDet, 2020 - 2021.
• Bayerische Gleichstellungsförderung Stipendium zur Förderung der Chancengleichheit für Frauen in Forschung und Lehre - April 2019, University of Passau
• Überbrückungsförderung für exzellente Nachwuchswissenschaftlerinnen – March 2018, University of Passau
• Research Grant issued by the IC 1207 COST Action for a research visit at the Institute for Language and Speech Processing in Athens, Greece.
• Travel and research grants awarded in the scope of the PARSEME project, for meetings in Athens 2014, Valetta 2015, Dubrovnik 2016.
• Travel and research grant for the Summer School on Machine Translation in La Rochelle, 2016.

• Text Mining Project(MSc and BSc) – Summer 2017, Winter 2017, Summer 2018, Summer 2019, University of Passau
• Information Retrieval and NLP (Msc and Bsc) – Winter 2018,Winter 2019 University of Passau
• Advanced Data Science Seminar (Msc and Bsc) - Summer 2021, University of Passau
• Machine Learnig Seminar - since Summer 2020 - a monthly seminar co-organized with chairs of Data Science, Applied Machine learning, Chair of Digital Image Processing and the Institute of Software Systems in Technical Applications of Computer Science (FORWISS)
• Digital Text I and II (BSc) – Winter 2014, University of Belgrade
• Multimedia Project (BSc) – Summer 2013, Winter 2013, Summer 2014 - University of Belgrade
• Lexical Recognition in Natural Language Processing (Msc) – Summer 2014, University of Belgrade
• Seminar on Linguistic Linked Open Data – University of Belgrade - Summer 2014

INVITED TALKS AND TUTORIALS

• J. Mitrović, “Neural Language Models for Abusive Language Detection“, keynote speech at the French-German Summerschool on AI with Industry 2021 organized by the University of Passau, Ecole Normale Supérieure Paris-Saclay and with the support from Siemens AI Lab.
• J. Mitrović, “Hate Speech Detection - not an easy Feat!“, invited talk at 10 Minuten Soziologie zum Thema Stress, University of Passau.
• J. Mitrović, “Computational rhetoric in Social Media – have we all become rhetoricians?”, invited lecture at the Waterloo.ai Institute Seminar, University of Waterloo, Canada, March 2020 (postponed due to COVID-19).
• J. Mitrović, “Abusive Language Detection in Social Media”, invited talk at the Passau Data Science Summit, PasDas, April 2020 (postponed due to COVID-19).
• J. Mitrović, “Python for Natural Language Processing”, invited talk at the PyCon conference, the 1st Balkan conference dedicated to Python programming, Belgrade, Serbia, November 18th, 2018.
• J. Mitrović, “Ontologies for the Humanities”, invited tutorial at the University of Waterloo, Canada, November 12th, 2018.
• J. Mitrović, “Building ontologies of rhetorical figures”, Computing Figures, Figuring Computers III – A Workshop in Computational Rhetoric”, University of Waterloo, Canada, invited tutorial, June 2018.
• J. Mitrović, “Ontology of Rhetorical Figures and its application to MWE research”, School of Computer Science, University of Valletta, Malta, invited talk, March 2015.
• J. Mitrović, “Linguistic Linked Open Data Paradigm”, Seminar of the Society for Language Technologies and Resources, Faculty of Mathematics, University of Belgrade, Serbia, May, 2015.
• J. Mitrović, “Processing of Rhetorical Figures for Serbian – Tools, Lexical resources and Implementation”, Computing Figures, Figuring Computers II – A Workshop in Computational Rhetoric”, University of Waterloo, Canada, invited talk, August 2016.

• PI in the University of Passau Project "Deep Write" in the scope of the KI Campus funding scheme.
• Member of the Board of Scientific Directors of the Rhetoricon project (University of Waterloo, Canada).
• Management Committe Member of the COST Action Nexus Linguarum.
• Organizer of the Abusive language dataset annotation workshop in Skopje, North Macedonia, as part of NexusLinguarum COST Action.
• Member of the Association for Computational Linguistics (ACL).
• ACL SIGLEX-MWE Standing Committee nominated officer, since September 2018.
• Founding Member of the Language Resources and Technology Group, University of Belgrade,Serbia.
• Quality reviewer for EU Project deliverables of the SSIX project https://ssix-project.eu/public-deliverables/
• Quality reviewer of the Max Planck Institute for Innovation and Competition working document about Artificial Intelligence and Copyright law.
• Program Committee Member at KEYSTONE 2015, KEYSTONE 2017, ACL 2019, ACL 2020, EMNLP-IJCNLP 2019, NAACL-HLT 2019, 34th AAAI Conference on Artificial Intelligence 2020, Natural Legal Language Processing Workshop 2020, AACL-IJCNLP 2020, EMNLP 2020, SemEval 2020, OffensEval 2020, 35th AAAI Conference on Artificial Intelligence 2021, IEEE CIC 2021, NLLP Workshop 2021, HASOC @ FIRE 2021, ACL Rolling Review 2022.
• Organizer and Program Committee chair of the MWE-WN Workshop at ACL 2019, and MWE-LEX at COLING 2020.
• Senior project supervisor at the Basis Independent Silicon Valley School, California, USA.
• Reviewer for the Journal of Natural Language Engineering published by Cambridge University Press, Infotheca Journal for Digital Humanities, and PLOS ONE.

• German Legal Citation Network
• HateBERT - retrained BERT language model for hate speech detection in English. Easy implementation now possible via the HuggingFace Transformers Library
• Our C-biGRU Offensive Language Detector detects offensive language in German and English, and is being updated for Hindi, Danish, Turkish and Serbian.
• Our Abusive language detection corpus, AbuseEval and the corresponding LREC 2020 paper are available on the GitHub page.
• German legal jugements annotated with judement style components
• German Legal Decisions Corpus