László János Laki
Personal
| Send me | |
| Phone | +36-1-886-4750 |
| About | I completed my PhD degree at the Multidisciplinary Technical Sciences Doctoral School at Pázmány Péter Catholic University in 2015. From 2012 I work in the MTA-PPKE Hungarian Language Technologies Research Group leaded by Prof. Dr. Gábor Prószéky. In NLPG I am working on some of the modules of a psycho-linguistically motivated text parser. During my doctoral studies I gained deeper experiments in the field of machine translation. In my work I increased the quality of the SMT translation in the case of morphologically rich languages. During the examination of the applicability of the SMT in additional areas I developed an SMT-based Part-of-Speech tagger, called HuLaPos. Besides my actual tasks in NLPG group I am also interested in deep learning, neural machine translation systems and quality estimation methods. |
| Research Interests | Natural Language Processing, Machine Translation, Quality estimation for MT, Translation Memories, Data mining, Machine Learning, Part-of-Speech tagging, Text annotation, Deep Learning |
EXPERIENCE
| Researcher, 2012 – |
MTA-PPKE Hungarian Language Technology Research Group, Budapest, Hungary Topic: We are working on a psycholinguistically motivated left-to-right text parser, called AnaGramma. |
| Natural Language Processing Engineer 2012 – |
MorphoLogic Localisation Ltd Topic: Our product (Globalese) is a machine translation system, which can be fully integrated into the translation workflow with the most popular professional Computer Aided Translation (CAT) tools. We have implemented the state-of-the-art machine translation techniques. |
Education
| Ph.D., 2009 – 2015 |
Pázmány Péter Catholic University Multidisciplinary Doctoral School of Sciences and Technology, Budapest, Hungary Thesis Topic: Efficient and high-quality solutions for mono- and multilingual issues based on methods of statistical machine translation |
| Postgraduate specialist training, 2009 – 2011 |
Pázmány Péter Catholic University, Computer Engineer specialized in Electronic Business, Budapest, Hungary Topic: Online catalogs Supervisor: István Vető |
| M.Sc., 2004 – 2009 |
Pázmány Péter Catholic University Faculty of Information Technology, Budapest, Hungary Topic: Morpheme-based Statistical Machine Translation System with Word Reordering Supervisor: Gábor Prószéky, DSc |
| Erasmus exchange student, Feb – Jul 2008 | Catholic University of Leuven, Leuven, Belgium |
| Summer School – September 2012 |
7th Machine Translation Marathon University of Edinburgh, Edinburgh, Scotland |
Teaching Experience
| Teaching Assistant |
Introduction to Programming I., II. Pázmány Péter Catholic University |
|
Introduction to Programming I., II. Pázmány Péter Catholic University |
Personal Skills
| Programming languages |
Expert in C++, Python, XML, GIT, BASH, AWK, XSLT Some experience with JAVA, PHP, HTML, MySQL, C#, C, Perl |
| Tools |
Expert in Eclipse framework, Code::Blocks, MS Word, MS Excel, MS PowerPoint, LATEX, PyCharmMS Windows, Ubuntu, Centos, Scrum framework Some experience with SWING widget, NLTK, OpenNLP MT specific tools: Moses, Nematus, Moses4Localization, Okapi/Tikal, Memoq, SDL Trados, QuEst |
| Endorsements | Computer Science, Natural Language Processing, Artificial Intelligence, Machine Translation, Big Data |
| Languages |
Hungarian (native) English (B2, intermediate exam, fluent) French (B2, intermediate exam) |
Publications
.
2020. Automatic Tag Recommendation for News Articles. 11th International Conference on Applied Informatics (ICAI 2020). :442-451.
(1.35 MB)
(1.35 MB) .
2020. Automatic Diacritic Restoration With Transformer Model Based Neural Machine Translation for East-Central European Languages. 11th International Conference on Applied Informatics (ICAI 2020). :190-202. (1.09 MB)
(1.09 MB) .
2020. Automatikus ékezetvisszaállítás transzformer modellen alapuló neurális gépi fordítással. XVI. Magyar SzámítógépesNyelvészeti Konferencia. :181-190. (292.27 KB)
(292.27 KB) .
2020. Automatikus összefoglaló generálás magyarnyelvre BERT modellel. XVI. Magyar Számítógépes Nyelvészeti Konferencia. :343-354. (344.42 KB)
(344.42 KB) .
2020. Automatikus tematikuscímke-ajánló rendszer sajtószövegekhez. XVI. Magyar Számítógépes Nyelvészeti Konferencia. :155-168. (1.48 MB)
(1.48 MB) .
2020. CBOW-tag: a Modified CBOW Algorithm for Generating Embedding Models from Annotated Corpora. Proceedings of the 12th Language Resources and Evaluation Conference. (337.69 KB)
(337.69 KB) .
2019. Creation of a corpus with semantic role labels for Hungarian. Proceedings of the 13th Linguistic Annotation Workshop. (273.57 KB)
(273.57 KB) .
2019. Egy magyar nyelvű kérdezőrendszer. XV. Magyar Számítógépes Nyelvészeti Konferencia. (845.56 KB)
(845.56 KB) .
2019. CBOW/A: módosított CBOW algoritmus annotált szövegekből készített vektortérmodellek létrehozására. XV. Magyar Számítógépes Nyelvészeti Konferencia. (1.55 MB)
(1.55 MB) .
2017. πRate: A Task-oriented Monolingual Quality Estimation System. International Journal of Computational Linguistics and Applications. (238.91 KB)
(238.91 KB) .
2017. Combining Machine Translation Systems with Quality Estimation. 18th International Conference on Intelligent Text Processing and Computational Linguistics (CICLing2017), Budapest, Hungary. (199.64 KB)
(199.64 KB) .
2017. Minőségbecslő rendszer egynyelvű természetes nyelvi elemzőhöz. XIII. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2015). :37–46.
.
2016. Mozaik nyelvmodell az AnaGramma elemzőhöz. XII. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2016). :260–270.
.
2016. HuQ: An English-Hungarian Corpus for Quality Estimation. Proceedings of the Tenth International Conference on Language Resources and Evaluation (LREC 2016). (152.2 KB)
(152.2 KB) .
2016. Quality Estimation for English-Hungarian with Optimized Semantic Features and the HuQ corpus. Computational Linguistics and Intelligent Text Processing - 17th International Conference, CICLing 2016, Konya, Turkey, April 3-9, 2016.
.
2016. Gépi fordítás minőségbecslésének optimalizálása kétnyelvű szótár és WordNet segítségével. XII. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2016). :37–46.
.
2015. Quality Estimation for English-Hungarian Machine Translation. 7th Language & Technology Conference. :170–174.
.
2015. Gépi fordítás minőségének becslése referencia nélküli módszerrel. XI. Magyar Számítógépes Nyelvészeti Konferencia (MSZNY 2015). :3–13.
.
2014. HuLaPos2 – Fordítsunk morfológiát. X. Magyar Szám{ítógépes Nyelvészeti Konferencia.
.
2013. Improved Hungarian Morphological Disambiguation with Tagger Combination. Text, Speech, and Dialogue. 8082:280–287.
.
2013. Combining Language-Independent Part-of-Speech Tagging Tools. 2nd Symposium on Languages, Applications and Technologies.
.
2013. Morfológiai egyértelműsítés nyelvfüggetlen annotáló módszerek kombinálásával. IX. Magyar Szám{ítógépes Nyelvészeti Konferencia. :331–337.
.
2013. Hunglish mondattan – átrendezésalapú angol-magyar statisztikai gépifordító-rendszer. IX. Magyar Szám{ítógépes Nyelvészeti Konferencia. :71–82.
.
2013. English to Hungarian Morpheme-based Statistical Machine Translation System with Reordering Rules. Proceedings of the Second ACL 2013 Workshop on Hybrid Approaches to Machine Translation (HyTra). :42-50.
.
2013. Syntax Based Reordering in Phrase Based English-Hungarian Statistical Machine Translation. International Journal of Computational Linguistics and Applications. 4:63–78.
.
2013. HuLaPos 2.0 – Decoding morphology. MICAI 2013: 12th Mexican International Conference on Artificial Intelligence. 8265:294–305.
.
2012. Investigating the Possibilities of Using SMT for Text Annotation. SLATE 2012 – Symposium on Languages, Applications and Technologies.