Anaïs Tack

BAEF Fellow - Postdoctoral Scholar
Department of Computer Science | Piech Lab
Stanford University

I'm a postdoctoral researcher interested in the application of natural language processing and machine learning in educational technologies for foreign language learners.

I obtained a BA and MA in English and French philology from KU Leuven. Subsequently, I studied language engineering and computer science at UCLouvain, where I graduated as Master in Computational Linguistics. I continued my graduate studies as an F.R.S.-FNRS research fellow at both UCLouvain and KU Leuven, where I obtained a joint Ph.D. degree in (Computational) Linguistics.
During my doctorate, I specialized in the automated prediction of lexical difficulty for foreign language readers. During this time, I was also involved in the ANR Alector project (2017-2020) on automatic text simplification for French children with reading disabilities and dyslexia. Before, I worked as a research assistant funded by the Altissia e-learning company on a project on the automated grading of short answers (ASAG) written by English learners with CEFR proficiency levels.
  • CEFR-ASAG: A corpus of short answers written by learners of English and graded with CEFR levels for CEFR-based short answer grading
  • NT2Lex: A CEFR-graded lexical resource for Dutch as a foreign language
  • Alector alignments: Alignments in the French Alector corpus with reading errors made by dyslexic and poor-reading children
I have been a PC member for several NLP conferences and workshops, including ACL, CoNLL, EACL, EMNLP, NAACL. I have also reviewed for CALL journals (CALICO, ReCALL) and was part of the OC for the EUROCALL 2019 conference.
I am currently a guest lecturer in NLP at UCLouvain (LFIAL2620) and give an invited lecture on NLP for CALL at KU Leuven (F0VH5A). I was a TA for several courses in NLP at UCLouvain (LCLIG2220, LINGI2263) and for French Proficiency I at KU Leuven Campus Kortrijk (V0AG3A).

Selected publications


Mark My Words!

On the Automated Prediction of Lexical Difficulty for Foreign Language Readers

UCLouvain KU Leuven


A CEFR-Graded Lexical Resource for Dutch as a Foreign Language Linked to Open Dutch WordNet

Tack Anaïs · François Thomas · Desmet Piet · Fairon Cédrick


Human and Automated CEFR-Based Grading of Short Answers

Tack Anaïs · François Thomas · Roekhaut Sophie · Fairon Cédrick


Modèles adaptatifs pour prédire automatiquement la compétence lexicale d’un apprenant de français langue étrangère

Tack Anaïs · François Thomas · Ligozat Anne-Laure · Fairon Cédrick


Deep Learning Architecture for Complex Word Identification

De Hertog Dirk · Tack Anaïs


A Report on the Complex Word Identification Shared Task 2018

Yimam Seid · Biemann Chris · Malmasi Shervin · Paetzold Gustavo · Specia Lucia · Štajner Sanja · Tack Anaïs · Zampieri Marcos



A Parallel Corpus of Simplified French Texts with Alignments of Misreadings by Poor and Dyslexic Readers

Gala Nuria · Tack Anaïs · Javourey-Drevet Ludivine · François Thomas · Ziegler Johannes



Foreign languages

  • English
  • French
  • Dutch

Complexity and difficulty

  • lexical complexity
  • complex word identification
  • lexical difficulty prediction


  • individual differences
  • variational linguistics
  • relative linguistic complexity


  • automated essay grading
  • automated short answer grading
  • automated writing proficiency assessment


  • graded reading materials
  • word decoding and comprehension
  • perception and attention
  • reading difficulties and dyslexia

Proficiency assessment

  • Common European Framework of Reference (CEFR)

Natural language processing

  • lexical networks
  • WordNet
  • word-sense disambiguation

Artificial neural networks

  • incremental machine learning
  • word and character embeddings
  • transformers


  • Python
  • C
  • Java
  • Perl

Computer-assisted language learning

  • technologically enhanced input
  • attention-drawing techniques
  • electronic glosses

Statistical models

  • generalized linear mixed-effects models