People

Valtteri Angström — MS student.  Valtteri works with the TOSKA Group under the lead of Dr Matti Luukkainen.  Valtteri and TOSKA bring the latest in software engineering and front-end technologies to Revita — to create a friendly user experience for the learners and teachers.  He is working on his MS degree Computer Science.
Ilmari Kylliäinen's People finder profile
Ilmari Kylliäinen's Research portal profile

Jue Hou — PhD student.  Jue's research focuses on modeling learner proficiency and novel approaches to language modeling with deep-learning neural nets. He also works on scalability of Revita's back end, and on educational data science — to analyze and model the data collected from proof-of-concept studies with real learners.   
Jue Hou's People finder profile
Jue Hou's Research portal profile

Silja M. Huttunen — PhD (University of Helsinki).  Organizing user studies and experiments with teachers and learners.

Anisia Katinskaia — PhD Student.  Anisia works on all aspects of user studies involving Russian and Finnish.  Her research is on deep-learning language models for detection of grammatical errors.  Revita needs to identify "alternative-correct" answers — important settings in learning, where there is more than one correct answer.
Anisia Katinskaia's People finder profile
Anisia Katinskaia's Research Portal profile

Ilmari Kylliäinen — MS student.  Ilmari works with the TOSKA Group, under the direction of Dr Matti Luukkainen, to bring the latest developments in front-end technologies into Revita — to assure the best possible User Experience for the learners and teachers.  He is working on his MS degree in Language Technology.
Ilmari Kylliäinen's People finder profile
Ilmari Kylliäinen's Research portal profile

Anh-Duc Vu — PhD Student.  Duc works on Bayesian networks and deep learning modeling of the learner proficiency and progress, which is essential for the personalization aspect in the Revita Project — learner competency must be modeled as accurately as possible to provide suitable exercises and feedback.

Duc's MS Thesis (2020) on "Modeling knowledge states in language learning."

Project lead.  Associate Professor at the Department of Digital Humanities, University of Helsinki (UH).  He holds the Chair in Linguistic Inequalities and Translation Technologies at INEQ: The Helsinki Inequality Initiative.  His research group — the Language Learning Lab — works on a variety of topics: studying how language works, and how computers can better understand language.  His earlier research themes include analysis of news media, and modeling language evolution.   Recent focus is on AI support for language learning.  This has resulted in Revita, a system currently in use by learners and teachers at several universities.  The research focuses on support for learning of smaller, endangered languages, as well as for "majority" languages.
Roman Yangarber's People finder profile
Publications (Google Scholar)

Here are some of our main contributors: local, at the University of Helsinki (UH), and international.
Remote Studies — current work with groups of learners at other universities, collecting feedback from them and researching their progress.

  • Matti Luukkainen — University lecturer (UH).  Leader of the TOSKA Reseach Group.  Principal advisor on best practices for Software Engineering, Web development and User Experience.  Supervising Gert Adamson, Valtteri Angström, Atte Haarni, Esa Kemppainen, Jami Kousa, Ilmari Kylliäinen, Tero Tapio, and other TOSKAns making substantial contributions.
  • Lari Kotilainen — University lecturer (UH).  Principal advisor for Finnish linguistics and didactics.
  • Mihail Kopotev — Professor (Higher School of Economics, St Petersburg, Russia), University lecturer (University of Helsinki).  Principal advisor for Russian linguistics and didactics.
  • Elena China-Kohlemainen — Master of Arts (UH).  Support for Italian: development of learning topics and feedback for learners.  Elena's MA Thesis (2021):
    "Computer-Assisted Language Learning based on Authentic Texts: Applications to Italian"
  • Maria Lebedeva — (Pushkin Institute, Moscow).  Leader on longitudinal studies with teachers of Russian.
  • Jouko Ahvenainen — (London, UK).  Advisor on user-centric approaches in artificial intelligence and machine learning.
  • Vesna Hendry, Tommaso Furlan — (Udine, Italy).  Design Team: design of user interface and  improving user experience.
  • Mikhail Mikhailov — Professor (Translation Studies, Tampere University, Finland).  Studies with university-level learners of Finnish and Russian.
  • Nataliya Stoyanova — PhD, Researcher (State University of Milan, Italy).  Principal advisor for Russian linguistics and didactics: studies with university-level learners of Russian.
  • Laila Paracchini — PhD, Research Fellow, Department of Foreign Languages and Literatures (State University of Milan, Italy).  Principal advisor for Russian: studies with university-level learners of Russian.
  • Alsu Zakirova — MA student (Higher School of Economics, St Petersburg, Russia).  Support for Tatar: development of linguistic topics and feedback for learners.
  • Mikhail Zolotilin — MA student (UH).  Enhancements to linguistic resources for Russian.
  • Jouni Rostila — PhD, University lecturer (UH).  Principal advisor for German linguistics and didactics.
  • Juha Toropainen — Master of Arts (UH).  Development of learning topics and feedback for German.
  • Riku Kuusijärvi — MA student (UH).  Development of learning topics and feedback for German.

All of our colleagues and collaborators typically work on multiple problems and aspects of the Project at the same time, and contribute to many areas.  We list here some of them here:

  • Javad Nouri — MS Student.  Javad built all components of the first functioning Revita platform.  Initial focus on several endangered Finno-Ugric languages, Finnish, Russian.
  • Kim Salmi — MS Student.  Contributed to many front-end and back-end aspects of the early platform.
  • José María (Txema) Hoya Quecedo — MS Student.  The core back-bone for supporting cross-language handling of morphological analysis and for (shallow) syntactic analysis.  NLP topics: models of ambiguity and disambiguation.  Initial support for Romance languages (β).
  • Max Koppatz — MS Student.  Major improvements to the language-analysis layer and to exercise selection.  NLP work: deep-learning language modeling.  Language-specific work on Finnish, and initial support for Swedish and German (β).
  • Tomi Rikander — MS Student.  Front-end improvements and re-engineering.
  • Sardana Ivanova — PhD Student.  Support for Russian, Sakha (Yakut).
  • Gert Adamson — MS Student.  Work with the TOSKA group to create the new REACT-based front-end, and enhanced user experience.