Esperanto grammar made easy!
Do you find the Esperanto grammar rules tricky? Then Lingvohelpilo can help!
Lingvohelpilo
A spell- and grammar-checker for Esperanto – easy, efficient and adjustable!Lingvohelpilo not only finds orthographical errors and false friends. It also helps you to master the accusative and to choose the correct words, affixes and inflections.
In addition, Lingvohelpilo has a pedagogical aspect. Unlike other proofing tools, it “understands” your text, analyzing the linguistic structure of words and sentences. Doing so allows it to identify and classify individual error types and to offer pedagogical explanations and examples.
You can use Lingvohelpilo directly in its web interface, or as a plug-in for Word, Google Docs and Outlook. It doesn’t matter whether you are working on a laptop, tablet or smartphone.
Lingvohelpilo is free, with or without a login. Registering your own user account will, however, allow you to set personal preferences and to create your own user dictionary. You can, for instance, ignore unsafe error markings or choose to exclusively focus on accusative errors.
The new edition of Lingvohelpilo has been funded by ESF (Esperantic Studies Foundation) and is being realised using GrammarSoft’s proofing software.
About Lingvohelpilo
The Lingvohelpilo 2 project proposes a new, improved and modernized version of the Esperanto spell- and grammar-checker Lingvohelpilo, the original version of which was developed between 2007 and 2009. The new, enhanced version works not only in a web interface, but also directly with Word, Google Docs and Outlook, on various platforms and different devices. This is achieved by porting Lingvohelpilo to an existing professional spell- and grammar-checker platform (GrammarSoft’s RetMig) that shares Lingvohelpilo’s focus on pedagogical aspects such as error typing and explanations.In addition, Lingvohelpilo’s databases and lexica are given a complete overhaul based on new language data harvested from the internet and other sources, and its analysis and error detection rules are being corrected and appended in order to achieve increased error coverage and to reduce the number of false positive alerts.
The Project
The project has a time frame of 15 months and will be concluded with an evaluation phase, where further improvements will be made based on user feedback, both in terms of performance and user-friendliness. The Lingvohelpilo-2 project is financed by ESF (Esperantic Studies Foundation) and carried out by the Danish company GrammarSoft.Project phases
- January – March 2024: Data collection and corpus compilation and annotation
- January – June 2024: Programming for porting Lingvohelpilo to the GrammarSoft proofing platform (RetMig) and integrating it into the new infrastructure
- March 2024 – May 2024: Main lexicography phase and data integration into the spellchecker
- May 2024 – December 2024: Main Constraint Grammar phase: Rules for morphosyntactic parsing and disambiguation, error detection rules, real word errors
- July 2024 – December 2024: Esperanto/Lingvohelpilo-specific frontend adaptations: JavaScript programming, pedagogical texts, interactive design
- Early 2025: Testing and feedback workshop, evaluation
- January – March 2025: Second Constraint Grammar phase: Correction and addition of rules based on feedback and evaluation, reprogramming of the graphical user interface (GUI) based on testing and user feedback
- 2025: Dissemination: Participation in one or two scientific conferences
Participants
Work on the new Lingvohelpilo is being carried out by Eckhard Bick and Tino Didriksen, using the spell- and grammar checking infrastructure of a small Danish language-technology company, GrammarSoft. Eckhard Bick is taking care of the linguistic tools proper, lexicography, corpus work, rule writing, pedagogical texts and design issues, as well as administration, while Tino Didriksen is handling front end (GUI) programming tasks, data crawling and the integration with third party software such as Word.Both Eckhard and Tino also participated in the original Lingvohelpilo project and have maintained the system since, with occasional backend and frontend adjustments to keep the service running on a university server (University of Southern Denmark).
External collaborators: For its later stages and future updates, the project actively seeks and welcomes community support for data collection and testing.
CVs
Dr. phil. Eckhard Bick is a computational linguist with a professional background in both Medicine and the Humanities. From 1996 to 2023 he has been affiliated with the Institute of Language and Communication at the University of Southern Denmark. In 2000 he co-founded GrammarSoft, a small language-technology company focusing on the grammatical annotation of running text and user applications linked to this task. Eckhard Bick has written computational grammars and lexica for most Germanic and Romance languages and is a leading expert in the area of Constraint Grammar. He has published extensively within the area of Corpus Linguistics and Natural Language Processing. Current research interests include semantic corpus annotation, social media, machine translation and rule-based proofing tools. Eckhard Bick has a long-standing engagement with Esperanto, including both teaching and research. Thus, he has published teaching materials (Tesi la testudo), dictionaries (e.g. Esperanto-Dana Vortaro), as well as other original and translated work in Esperanto, and participated in many Esperanto projects, among them Lingvohelpilo and two EU-funded Esperanto projects, Lingvo.info and Multilingualism Accelerator.
Tino Didriksen is an independent programmer and software consultant with a strong knowledge of C, C++, Java, Perl, PHP, JavaScript, HTML and XML, as well as server setup and maintenance on Linux platforms. He has worked with Eckhard Bick both on Southern Denmark University’s VISL project (since 2004) and for GrammarSoft (since 2007), providing support for a large variety of tasks, including GrammarSoft’s proofing tools and several Esperanto projects, among them the original Lingvohelpilo (2007-2009). In addition, he works for Learn Greenlandic and the Greenlandic Language Secretariat (Oqaasileriffik), and maintains his own chat site, pJJ Chats.
Collaborate
We welcome all the help we can get, covering areas of expertise such as the following:- teachers who can compile information about frequent and/or typical errors, false friends etc.
- linguists/grammarians who can suggest rules for identifying and correcting grammatical errors
- students of (computational) linguistics who can carry out supplementary research on Esperanto topics
- language-minded programmers
- translators who can translate Lingvohelpilo’s pedagogical comments and meta texts into further languages
Collecting errors
Lingvohelpilo depends on human-crafted linguistic rules that identify and correct the very diverse kinds of errors made by Esperanto speakers. In order to write good rules, it is important to have access to a large and varied collection of error examples. Help with collection such data would be very much appreciated.If you decide to help with this, please make sure to include the following types of information:
- the entire sentence containing the error, to provide necessary context
- the error itself and a correction proposal
- how would you classify the error? Spelling, grammar, word choice? False friends? Neologisms?
- optionally information about the source: internet, journal, social network etc.? Learner or fluent speaker?
Pedagogical comments
Under preferences you can find a list of error types. If you know, or can create, a good longer text about a given topic, with explanations, examples or even exercises, we can link to this text as additional material, on top of the existing, shorter texts. Thus, in may places, we are already using links to the teaching portal Lernu!Testing
Although the new Lingvohelpilo is by no means finished yet, it is already possible to test various aspects and its basic functionality. If you come across texts with errors, try to copy the sentence(s) in question and check it out in Lingvohelpilo’s online interface, noting if the program can handle the error or not. If Lingvohelpilo offers explanations, note whether you found them helpful. Towards the end of the project, it will also be possible to test for technical problems related to operating systems, browser or the integration with Word and Google Docs.Remember that the main target group of Lingvohelpilo are intermediate and advanced language learners, or in general people who do not make more than 1-2 errors per sentence. A higher error density will affect performance because the context of a to-be-identified error may itself contain errors and thus become unreliable.
Financial support
In order to ensure and maintain optimal performance and continued development, finances are important. You can make an ear-marked donation through ESF (Esperantic Studies Foundation) who are bankrolling the project.Thanks for helping!
Further information
What is the difference between spell- and grammar-checking?
Spellcheckers typically compare words with an official word list. They can flag unknown words as likely errors, but will not complain about real word errors, i.e. words that do exist, but are wrong in a given context. Grammar checkers, on the other hand, also recognize errors in existing words, e.g. the confusion of -i and -as, or the lack of an accusative inflection.How does the automatic text analysis in Lingvohelpilo work?
Both for the underlying linguistic analysis and for the subsequent mark-up of grammatical errors we employ so-called CG rules (Constraint Grammar). The method exploits the sentence context, together with lexical and morphological information, to assign tags for word class, inflection, syntactic function, semantic category etc. For the linguistic analysis, the EspGram parser is used, providing input for Lingvohelpilo itself.Which texts were used during the creation of Lingvohelpilo?
We used the text collections of CorpusEye, which were expanded and re-annotated in connection with the Lingvohelpilo project. Temas entute pri pli ol 210 milionoj de vortoj, in four main sections:• Classical and translated literature
• Journals (Revuo Esperanto, Monato, Kontakto, tejo-tutmonde, El Popola Ĉinio, Eventoj)
• Wikipedia
• Crawled web corpus, blogs etc.
It is possible to search these text collections through CorpusEye’s graphical interface, find usage examples for teaching or perform various statistics, comparisons and visualisations for linguistic research.
Where can I find scientific publications about EspGram, Constraint Grammar, the text corpora and the error detection methods used in Lingvohelpilo?
You can find a number of relevant publications about Esperanto in the list of VISL-publications, as well as general papers about Constraint Grammar and other languages using the same methods, e.g.:- Bick, Eckhard (2023). VISL & CG-3: Constraint Grammar on the Move: An application-driven paradigm. In: Arvi Hurskainen, Kimmo Koskenniemi & Tommi Pirinen (eds.), Rule-Based Language Technology. NEALT Monograph Series vol. 2, pp. 112-140. University of Tartu. ISSN 1736-6291
- Bick, Eckhard (2020). Syntax and Semantics in a Treebank for Esperanto. In: Calzolari, Nicoletta et al. (eds.), Proceedings of the 12th International Conference on Language Resources and Evaluation, LREC2020 (Marseille, May 2020). pp. 5122-5129. ACL / ELRA
- Bick, Eckhard (2019). Automatic Generation and Semantic Grading of Esperanto Sentences in a Teaching Context. In: Alfter, David et al. (eds.) Proceedings of the 8th Workshop on NLP for Computer-Assisted Language Learning (NLP2CALL at NoDaLiDa, Sept. 2019, Turku, Finland). NEALT Proceedings Series Vol. 39 & ACL Anthology W19-6302, pp. 10-19. LiU Electronic Press & ACL. ISBN 978-91-7929-998-9
- Bick, Eckhard (2018). Arbobanko - A Treebank for Esperanto. In: Proceedings of CICLing 2018 - 19th International Conference on Computational Linguistics (Hanoi, March 2018), forthcoming (preprint, LNCS - Lecture Notes in Computer Science, Springer)
- Bick, Eckhard (2016). A Morphological Lexicon of Esperanto with Morpheme Frequencies, In: Calzolari, Nicoletta et al. (eds.), Proceedings of the 10th International Conference on Language Resources and Evaluation, LREC2016 (Portorož, May 23-28, 2016). pp. 1075-1078. ISBN 978-2-9517408-9-1
- Bick, Eckhard (2015). DanProof: Pedagogical Spell and Grammar Checking for Danish. In: Galia Angelova, Kalina Bontcheva & Ruslan Mitkov: Proceedings of RANLP 2015 (Hissar, Bulgaria, 7-9 Sept. 2015). pp. 55-62. ISSN 1313-8502, ISBN 781510813281
- Bick, Eckhard (2015). La kunteksta vortaro DeepDict: Korpusa leksikografio en Esperanto. In: Christer Kiselman & José Vergara (eds.): De esperanta korpuso ĝis islanda lingvopolitiko - Aktoj de la 36-a Esperantologia Konferenco en la 98-a Universala Kongreso de Esperanto, Rejkjaviko 2013. pp. 7-17. Rotterdam: Universala Esperanto-Asocio. ISBN 978-92-9017-125-6
- Bick, Eckhard (2011). WikiTrans, The English Wikipedia in Esperanto. In: Constraint Grammar Applications, Workshop at NODALIDA 2011, Riga, Latvia. (forthcoming: NEALT Proceedings Series, Vol. 14, pp. 8-16. Tartu: Tartu University Library. ISSN 1736-6305)
- Bick, Eckhard (2011), WikiTrans: La angla Vikipedio en Esperanto. In: Modernaj teknologioj por Esperanto (Proceedings of KAEST 2010, Modra, Slovakia). pp. 28-41. Espero/E@I, Partizánske. ISBN 978-80-89366-10-1
- Bick, Eckhard (2011), DeepDict - Konteksta reta vortaro de vera lingvouzo, In: Modernaj teknologioj por Esperanto (Proceedings of KAEST 2010, Modra, Slovakia). pp. 15-27. Espero/E@I, Partizánske. ISBN 978-80-89366-10-1
- Bick, Eckhard (2009). A Dependency Constraint Grammar for Esperanto. Constraint Grammar Workshop at NODALIDA 2009, Odense. NEALT Proceedings Series, Vol 8, pp.8-12. Tartu: Tartu University Library. (ISSN 1736-6305)
Contact us
Eckhard Bick (linguistic and pedagogical questions, corpora, research, publications) [Esperanto or English]Tino Didriksen (technical questions, programming, user interface) [English]
Lingvohelpilo is freely usable both with and without a user account. If you log in with an account, you will have access to the options and personal dictionary.
Continue without login
Create account Log in
Continue without login
Create account Log in