---
title: "Estonia — research institutions, scientists and papers"
description: "Estonia: 2 institutions with papers in our library in 2 agglomerations, 3 people with a page, 12 works, 64,851 citations."
canonical: "https://research.elamur.ai/free/en/countries/estonia/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/countries/estonia/)

[home](https://md.research.elamur.ai/free/en/) / [countries](https://md.research.elamur.ai/free/en/countries/) / Estonia

# Estonia

Country 2 institutions in 2 agglomerations · [all countries](https://md.research.elamur.ai/free/en/countries/)

[**2** institutions with a page](https://md.research.elamur.ai/free/en/countries/estonia/#institutions) · [**2** agglomerations](https://md.research.elamur.ai/free/en/countries/estonia/#agglomerations) · [**3** people with a page](https://md.research.elamur.ai/free/en/countries/estonia/#authors) · [**12** works in our library](https://md.research.elamur.ai/free/en/countries/estonia/#works) · [**64,851** citations of their works](https://md.research.elamur.ai/free/en/countries/estonia/#works)

## About

### Profile

Estonia appears on this site through two universities in two cities. [Tallinn University of Technology](https://md.research.elamur.ai/free/en/universities/tallinn-university-of-technology/) is in [Tallinn](https://md.research.elamur.ai/free/en/agglomerations/tallinn/), and the [University of Tartu](https://md.research.elamur.ai/free/en/universities/university-of-tartu/) is in [Tartu](https://md.research.elamur.ai/free/en/agglomerations/tartu/). Both are universities. No research institutes, hospitals or companies from the country have works here yet. The country can be compared with others through the [index of countries](https://md.research.elamur.ai/free/en/countries/). Its institutions and urban areas each have an overview of their own.

Among the works here, the country's research falls into two clusters. The first is open-source scientific software and computational reasoning, linked to Tallinn. The second is machine translation and its evaluation, centred on Tartu. A single study of ancient biblical translation adds a humanities strand to the Tartu collection. The Tartu group is larger in the number of works and people on this site. Tallinn draws more citations, because two of the most widely used software papers in scientific computing are associated with it.

### History

On this site, Estonian research begins in 2013 with a paper on computational morphology co-written by [Kairit Sirts](https://md.research.elamur.ai/free/en/authors/kairit-sirts/). The next period starts in 2017, when University of Tartu authors begin to appear on the yearly findings reports of the Conference on Machine Translation (WMT). Those reports continue in 2018 and 2019. In 2019 they were joined by a findings report on the quality estimation shared task. In 2020 came a research article on estimating translation quality without reference translations.

Also in 2019 and 2020, Tallinn University of Technology is associated with the papers describing SciPy 1.0 and NumPy. These are the reference publications for two foundational Python libraries. The SciPy paper notes that the library had been in use since an early version released in 2001. The Estonian involvement shown here, however, dates from the 2019 publication.

After a gap, the machine translation line resumes with the WMT general task reports of 2024 and 2025. These follow the field's move to systems built on large language models. In 2024 a study of the Syriac Peshitta Psalms also appears under the University of Tartu. In 2026 a Tallinn paper on abstract visual reasoning extends the country's presence into the evaluation of reasoning in learned models.

### Research profile

The strongest theme among the works here is [machine translation](https://md.research.elamur.ai/free/en/topics/machine-translation/). It also forms its own [machine translation shelf](https://md.research.elamur.ai/free/en/shelves/machine-translation/) and sits within the broader field of [linguistics and languages](https://md.research.elamur.ai/free/en/shelves/linguistics/). Closely tied to it is the topic of [evaluation and benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/). Most of the Tartu translation papers are shared-task reports. These are large collaborative documents that set common test sets, gather system submissions and compare them with automatic metrics and human judgement.

The series includes [Findings of the 2017 Conference on Machine Translation (WMT17)](https://md.research.elamur.ai/free/en/papers/bojar-2017-findings-2017-conference-machine-translation/), which describes the annual campaign that follows progress in the field and provides shared test material. It continues with [Findings of WMT18](https://md.research.elamur.ai/free/en/papers/bojar-2018-findings-2018-conference-machine-translation/), the third edition of the conference, and [Findings of WMT19](https://md.research.elamur.ai/free/en/papers/barrault-2019-findings-2019-conference-machine-translation/). Later volumes are [the WMT24 general task report](https://md.research.elamur.ai/free/en/papers/kocmi-2024-findings-wmt24-general-machine-translation/) and [the WMT25 general task report](https://md.research.elamur.ai/free/en/papers/kocmi-2025-findings-wmt25-general-machine-translation/). Their subtitles summarise their arguments: that the arrival of large language models has not made translation a solved problem, and that evaluation must move beyond easy test sets.

A second strand within translation research is quality estimation, which means predicting how good a machine translation is without a human reference. [Findings of the WMT 2019 Shared Tasks on Quality Estimation](https://md.research.elamur.ai/free/en/papers/fonseca-2019-findings-wmt-2019-shared-tasks/) reports on the shared task devoted to this problem. [Unsupervised Quality Estimation for Neural Machine Translation](https://md.research.elamur.ai/free/en/papers/fomicheva-2020-unsupervised-quality-estimation-neural-machine/) proposes ways to judge output quality from the translation system itself. It targets languages that lack labelled training material, and it releases a resource covering both black-box and glass-box settings.

[Natural language processing](https://md.research.elamur.ai/free/en/shelves/nlp/) beyond translation is represented by work on [morphology](https://md.research.elamur.ai/free/en/topics/morphology/). [Minimally-Supervised Morphological Segmentation using Adaptor Grammars](https://md.research.elamur.ai/free/en/papers/sirts-2013-minimally-supervised-morphological-segmentation-adaptor/) studies how to split words into their meaningful parts with little labelled input. It compares unsupervised adaptor grammars built on hand-designed grammars with settings that add small amounts of supervision.

[Mathematics and statistics](https://md.research.elamur.ai/free/en/shelves/mathematics/) enters through scientific computing. [SciPy 1.0: Fundamental Algorithms for Scientific Computing in Python](https://md.research.elamur.ai/free/en/papers/virtanen-2019-scipy-1-0-fundamental-algorithms/) and [Array Programming with NumPy](https://md.research.elamur.ai/free/en/papers/harris-2020-array-programming-numpy/) describe two open-source libraries. NumPy provides the core array structure for numerical work in Python. SciPy builds on it with routines for optimisation, integration, linear algebra, statistics and related tasks.

The topic of [reasoning](https://md.research.elamur.ai/free/en/topics/reasoning/) appears in [Symbolic Grounding Reveals Representational Bottlenecks in Abstract Visual Reasoning](https://md.research.elamur.ai/free/en/papers/vaishnav-2026-symbolic-grounding-reveals-representational-bottlenecks/), by Mohit Vaishnav and Tanel Tammet. The authors treat their symbolically grounded configuration not as a system for everyday use but as a diagnostic ceiling. Its aim is to locate where representations limit performance on abstract visual puzzles.

A humanities line covers [Semitic languages](https://md.research.elamur.ai/free/en/topics/semitic-languages/) and [Hebrew and Aramaic](https://md.research.elamur.ai/free/en/topics/hebrew-aramaic/). Its only work here is [Psalm 29 as a poetological example of Peshitta Psalms translation](https://md.research.elamur.ai/free/en/papers/vasheghanifarahani-2024-psalm-29-poetological-example-peshitta/) by Amir Vasheghanifarahani. It takes one psalm as a case study in how the Syriac translators handled the poetry of their source, and it addresses a neglected question in Peshitta scholarship. It is also the second Estonian line here about translation, this time historical translation by human hands.

### Leading scientists

[Mark Fishel](https://md.research.elamur.ai/free/en/authors/mark-fishel/) of the University of Tartu comes first among the country's people on this site. Fishel's work here runs through the machine translation evaluation campaigns. The name appears among the authors of the WMT18 findings report, part of the WMT series that on this site spans 2017 to 2025. These reports compare translation systems across many language pairs in shared evaluations.

[Lisa Yankovskaya](https://md.research.elamur.ai/free/en/authors/lisa-yankovskaya/), also at the University of Tartu, works on quality estimation. Yankovskaya is a co-author of the 2019 quality estimation findings report and of the 2020 paper on unsupervised quality estimation. Both deal with judging machine translation output when no human reference is available.

Kairit Sirts is connected with both Tallinn University of Technology and the University of Tartu. Sirts's earliest work here is the 2013 study of morphological segmentation with adaptor grammars, written with Sharon Goldwater. It belongs to a line of research on learning the structure of words from mostly unlabelled text, which matters for languages with rich word forms.

[Krister Lindén](https://md.research.elamur.ai/free/en/authors/krister-linden/) is also connected to the University of Tartu on this site.

### Landmark works

By citation, the most prominent works are the two software papers associated with Tallinn University of Technology. The 2019 SciPy 1.0 paper is the most cited of all. It explains the library's design, its algorithms and the community development behind it. The 2020 NumPy paper follows. It sets out the array programming model that a great deal of scientific Python code relies on. Both are large community efforts, with authors including Pauli Virtanen, Ralf Gommers and Travis E. Oliphant for SciPy, and Charles R. Harris, K. Jarrod Millman and Stéfan J. van der Walt for NumPy.

Next come the WMT findings reports from the University of Tartu. Of these, the WMT19 report led by Loïc Barrault, Ondřej Bojar and Marta R. Costa-jussà is cited most. The WMT18 and WMT17 reports, both led by Ondřej Bojar, follow closely. These reports serve as reference points for which systems and methods performed best in a given year. The 2024 general task report and the 2019 quality estimation report are also well cited. So is the 2020 unsupervised quality estimation paper by Marina Fomicheva, Shuo Sun, Lisa Yankovskaya and colleagues.

The 2013 morphology paper by Sirts and Goldwater has a smaller following. The 2025 WMT report, the 2026 reasoning study and the 2024 Peshitta study are the most recent works and have only begun to collect citations.

### Labs and units

On this site, Estonia's research is organised through its two universities. Tallinn University of Technology holds the scientific computing papers, the morphology study and the visual reasoning paper. The University of Tartu holds the machine translation reports, the quality estimation work and the Peshitta study. No departments or smaller units of either university are shown separately here.

### Collaborations

Almost every Estonian work here is collaborative, and many are multinational. The WMT reports bring Tartu authors together with organisers such as Ondřej Bojar, Christian Federmann, Rajen Chatterjee, Loïc Barrault and Marta R. Costa-jussà in the earlier volumes. The later volumes add Tom Kocmi, Eleftherios Avramidis, Rachel Bawden and Ekaterina Artemova. The quality estimation work links Lisa Yankovskaya with Erick Fonseca and André F. T. Martins, and with Marina Fomicheva and Shuo Sun.

At Tallinn, the morphology paper pairs Kairit Sirts with Sharon Goldwater. The reasoning paper is a two-author study by Mohit Vaishnav and Tanel Tammet. The SciPy and NumPy papers are among the broadest collaborations on the site, joining many contributors to open-source software.

Kairit Sirts is the one person with a page here connected to both universities. This makes Sirts a direct link between language technology in Tallinn and in Tartu.

### Timeline

In 2013 the country's first work here appears: the adaptor grammar study of morphological segmentation, co-written by Kairit Sirts.

From 2017 to 2019, University of Tartu authors take part in three successive WMT findings reports. In 2019 the quality estimation shared task report is added. In the same year Tallinn University of Technology is associated with the SciPy 1.0 paper.

In 2020 the NumPy paper appears, alongside the Tartu study of unsupervised quality estimation for neural translation.

In 2024 and 2025 the WMT general task reports return. They track translation in the era of large language models and argue for harder test sets. In 2024 Tartu also adds the study of Psalm 29 in the Peshitta.

In 2026 a Tallinn paper on symbolic grounding in abstract visual reasoning becomes the most recent Estonian work in our library.

## Where

## Agglomerations (2)

1. [**Tallinn**1 institution*62,003 citations*](https://md.research.elamur.ai/free/en/agglomerations/tallinn/)
2. [**Tartu**1 institution*2,848 citations*](https://md.research.elamur.ai/free/en/agglomerations/tartu/)

## Institutions (2)

### 🎓 Universities (2)

1. **[Tallinn University of Technology](https://md.research.elamur.ai/free/en/universities/tallinn-university-of-technology/)**University · [Tallinn](https://md.research.elamur.ai/free/en/agglomerations/tallinn/), Estonia*4 works · 62,003 citations*
2. **[University of Tartu](https://md.research.elamur.ai/free/en/universities/university-of-tartu/)**University · [Tartu](https://md.research.elamur.ai/free/en/agglomerations/tartu/), Estonia*8 works · 2,848 citations*

## People (3)

Authors with a page confirmed at its institutions, the most cited first.

1. [**Mark Fishel**University of Tartu*8 papers here · 2,498 citations*](https://md.research.elamur.ai/free/en/authors/mark-fishel/)
2. [**Lisa Yankovskaya**University of Tartu*2 papers here · 484 citations*](https://md.research.elamur.ai/free/en/authors/lisa-yankovskaya/)
3. [**Kairit Sirts**Tallinn University of Technology*3 papers here · 231 citations*](https://md.research.elamur.ai/free/en/authors/kairit-sirts/)

## Works in our library (12)

Works of its institutions, the most cited first.

1. [**SciPy 1.0 — Fundamental Algorithms for Scientific Computing in Python**Pauli Virtanen, Ralf Gommers, Travis E. Oliphant, et al.*2019 · 38,317 citations*](https://md.research.elamur.ai/free/en/papers/virtanen-2019-scipy-1-0-fundamental-algorithms/)
2. [**Array Programming with NumPy**Charles R. Harris, K. Jarrod Millman, Stéfan J. van der Walt, et al.*2020 · 23,591 citations*](https://md.research.elamur.ai/free/en/papers/harris-2020-array-programming-numpy/)
3. [**Findings of the 2019 Conference on Machine Translation (WMT19)**Loïc Barrault, Ondřej Bojar, Marta R. Costa-jussà, et al.*2019 · 839 citations*](https://md.research.elamur.ai/free/en/papers/barrault-2019-findings-2019-conference-machine-translation/)
4. [**Findings of the 2018 Conference on Machine Translation (WMT18)**Ondřej Bojar, Christian Federmann, Mark Fishel, et al.*2018 · 747 citations*](https://md.research.elamur.ai/free/en/papers/bojar-2018-findings-2018-conference-machine-translation/)
5. [**Findings of the 2017 Conference on Machine Translation (WMT17)**Ondřej Bojar, Rajen Chatterjee, Christian Federmann, et al.*2017 · 685 citations*](https://md.research.elamur.ai/free/en/papers/bojar-2017-findings-2017-conference-machine-translation/)
6. [**Findings of the WMT24 General Machine Translation Shared Task: The LLM Era Is Here but MT Is Not Solved Yet**Tom Kocmi, Eleftherios Avramidis, Rachel Bawden, et al.*2024 · 177 citations*](https://md.research.elamur.ai/free/en/papers/kocmi-2024-findings-wmt24-general-machine-translation/)
7. [**Findings of the WMT 2019 Shared Tasks on Quality Estimation**Erick Fonseca, Lisa Yankovskaya, André F. T. Martins, et al.*2019 · 173 citations*](https://md.research.elamur.ai/free/en/papers/fonseca-2019-findings-wmt-2019-shared-tasks/)
8. [**Unsupervised Quality Estimation for Neural Machine Translation**Marina Fomicheva, Shuo Sun, Lisa Yankovskaya, et al.*2020 · 153 citations*](https://md.research.elamur.ai/free/en/papers/fomicheva-2020-unsupervised-quality-estimation-neural-machine/)
9. [**Minimally-Supervised Morphological Segmentation using Adaptor Grammars**Kairit Sirts, Sharon Goldwater*2013 · 94 citations*](https://md.research.elamur.ai/free/en/papers/sirts-2013-minimally-supervised-morphological-segmentation-adaptor/)
10. [**Findings of the WMT25 General Machine Translation Shared Task: Time to Stop Evaluating on Easy Test Sets**Tom Kocmi, Ekaterina Artemova, Eleftherios Avramidis, et al.*2025 · 73 citations*](https://md.research.elamur.ai/free/en/papers/kocmi-2025-findings-wmt25-general-machine-translation/)
11. [**Symbolic Grounding Reveals Representational Bottlenecks in Abstract Visual Reasoning**Mohit Vaishnav, Tanel Tammet*2026 · 1 citation*](https://md.research.elamur.ai/free/en/papers/vaishnav-2026-symbolic-grounding-reveals-representational-bottlenecks/)
12. [**Psalm 29 as a poetological example of Peshitta Psalms translation**Amir Vasheghanifarahani*2024 · 1 citation*](https://md.research.elamur.ai/free/en/papers/vasheghanifarahani-2024-psalm-29-poetological-example-peshitta/)

## Concepts, shelves and topics

Of its works in our library: the number of works on each and its rating among them.

💡 Concepts 1[Subword tokenization · 1 work](https://md.research.elamur.ai/free/en/concepts/subword-tokenization/)

📚 Shelves 4[Mathematics & statistics · 2 works](https://md.research.elamur.ai/free/en/shelves/mathematics/) · [Machine translation · 7 works](https://md.research.elamur.ai/free/en/shelves/machine-translation/) · [Linguistics & languages · 8 works](https://md.research.elamur.ai/free/en/shelves/linguistics/) · [Natural language processing · 2 works](https://md.research.elamur.ai/free/en/shelves/nlp/)

🧭 Topics 6[Machine translation · 7 works](https://md.research.elamur.ai/free/en/topics/machine-translation/) · [Evaluation & benchmarks · 6 works](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · [Morphology · 1 work](https://md.research.elamur.ai/free/en/topics/morphology/) · [Reasoning · 1 work](https://md.research.elamur.ai/free/en/topics/reasoning/) · [Hebrew & Aramaic · 1 work](https://md.research.elamur.ai/free/en/topics/hebrew-aramaic/) · [Semitic languages · 1 work](https://md.research.elamur.ai/free/en/topics/semitic-languages/)

## Works by year

### 2020s

[**2026** 1 work](https://md.research.elamur.ai/free/en/papers/vaishnav-2026-symbolic-grounding-reveals-representational-bottlenecks/) · [**2025** 1 work](https://md.research.elamur.ai/free/en/papers/kocmi-2025-findings-wmt25-general-machine-translation/) · [**2024** 2 works](https://md.research.elamur.ai/free/en/countries/estonia/year/2024/) · [**2020** 2 works](https://md.research.elamur.ai/free/en/countries/estonia/year/2020/)

### 2010s

[**2019** 3 works](https://md.research.elamur.ai/free/en/countries/estonia/year/2019/) · [**2018** 1 work](https://md.research.elamur.ai/free/en/papers/bojar-2018-findings-2018-conference-machine-translation/) · [**2017** 1 work](https://md.research.elamur.ai/free/en/papers/bojar-2017-findings-2017-conference-machine-translation/) · [**2013** 1 work](https://md.research.elamur.ai/free/en/papers/sirts-2013-minimally-supervised-morphological-segmentation-adaptor/)

- [← back to](https://md.research.elamur.ai/free/en/countries/) — countries

Updated 23 October 2026