@inProceedings{lange-ljunglof-2020-learning-291243, title = {Learning Domain-specific Grammars from a Small Number of Examples}, abstract = {In this paper we investigate the problem of grammar inference from a different perspective. The common approach is to try to infer a grammar directly from example sentences, which either requires a large training set or suffers from bad accuracy. We instead view it as a problem of grammar restriction or sub-grammar extraction. We start from a large-scale resource grammar and a small number of examples, and find a sub-grammar that still covers all the examples. To do this we formulate the problem as a constraint satisfaction problem, and use an existing constraint solver to find the optimal grammar. We have made experiments with English, Finnish, German, Swedish and Spanish, which show that 10–20 examples are often sufficient to learn an interesting domain grammar. Possible applications include computer-assisted language learning, domain-specific dialogue systems, computer games, Q/A-systems, and others.}, booktitle = {12th International Conference on Agents and Artificial Intelligence - Volume 1: NLPinAI}, author = {Lange, Herbert and Ljunglöf, Peter}, year = {2020}, publisher = {SciTePress}, ISBN = {978-989-758-395-7}, } @inProceedings{lange-ljunglof-2018-demonstrating-274016, title = {Demonstrating the MUSTE Language Learning Environment}, abstract = {We present a language learning application that relies on grammars to model the learning outcome. Based on this concept we can provide a powerful framework for language learning exercises with an intuitive user interface and a high reliability. Currently the application aims to augment existing language classes and support students by improving the learner attitude and the general learning outcome. Extensions beyond that scope are promising and likely to be added in the future.}, booktitle = {NLP4CALL 2018, the 7th Workshop on NLP for Computer Assisted Language Learning, Stockholm, 7th November 2018; published as issue 152 of Linköping Electronic Conference Proceedings}, author = {Lange, Herbert and Ljunglöf, Peter}, year = {2018}, publisher = {Linköping University Electronic Press, Linköpings universitet}, address = {Linköping}, ISBN = {978-91-7685-173-9}, } @inProceedings{bamutura-etal-2020-towards-296511, title = {Towards Computational Resource Grammars for Runyankore and Rukiga}, abstract = {In this paper, we present computational resource grammars of Runyankore and Rukiga (R&R) languages. Runyankore and Rukiga are two under-resourced Bantu Languages spoken by about 6 million people indigenous to South Western Uganda, East Africa. We used Grammatical Framework (GF), a multilingual grammar formalism and a special-purpose functional programming language to formalise the descriptive grammar of these languages. To the best of our knowledge, these computational resource grammars are the first attempt to the creation of language resources for R&R. In Future Work, we plan to use these grammars to bootstrap the generation of other linguistic resources such as multilingual corpora that make use of data-driven approaches to natural language processing feasible. In the meantime, they can be used to build Computer-Assisted Language Learning (CALL) applications for these languages among others.}, booktitle = {Proceedings of The 12th Language Resources and Evaluation Conference}, author = {Bamutura, David and Ljunglöf, Peter and Nabende, Peter}, year = {2020}, publisher = {European Language Resources Association}, } @inProceedings{bamutura-ljunglof-2019-towards-284293, title = {Towards a resource grammar for Runyankore and Rukiga}, abstract = {Currently, there is a lack of computational grammar resources for many under-resourced languages which limits the ability to develop Natural Language Processing (NLP) tools and applications such as Multilingual Document Authoring, Computer-Assisted Language Learning (CALL) and Low-Coverage Machine Translation (MT) for these languages. In this paper, we present our attempt to formalise the grammar of two such languages: Runyankore and Rukiga. For this formalisation we use the Grammatical Framework (GF) and its Resource Grammar Library (GF-RGL).}, booktitle = {WiNLP 2019, the 3rd Workshop on Widening NLP, Florence, Italy, 28th July 2019}, author = {Bamutura, David and Ljunglöf, Peter}, year = {2019}, } @techreport{ljunglof-etal-2019-assessing-281222, title = {Assessing the quality of Språkbanken’s annotations}, abstract = {Most of the corpora in Språkbanken Text consist of unannotated plain text, such as almost all newspaper texts, social media texts, novels and official documents. We also have some corpora that are manually annotated in different ways, such as Talbanken (annotated for part-of-speech and syntactic structure), and the Stockholm Umeå Corpus (annotated for part-of-speech). Språkbanken’s annotation pipeline Sparv aims to automatise the work of automatically annotating all our corpora, while still keeping the manual annotations intact. When all corpora are annotated, they can be made available, e.g., in the corpus searh tools Korp and Strix. Until now there has not been any comprehensive overview of the annotation tools and models that Sparv has been using for the last eight years. Some of them have not been updated since the start, such as the part-of-speech tagger Hunpos and the dependency parser MaltParser. There are also annotation tools that we still have not included, such as a constituency-based parser. Therefore Språkbanken initiated a project with the aim of conducting such an overview. This document is the outcome of that project, and it contains descriptions of the types of manual and automatic annotations that we currently have in Språkbanken, as well as an incomplete overview of the state-of-the-art with regards to annotation tools and models. }, author = {Ljunglöf, Peter and Zechner, Niklas and Nieto Piña, Luis and Adesam, Yvonne and Borin, Lars}, year = {2019}, } @misc{alexandersson-etal-2015-proceedings-228416, title = {Proceedings of SLPAT-2015: 6th Workshop on Speech and Language Processing for Assistive Technologies}, abstract = {We are pleased to bring you the Proceedings of the 6th Workshop on Speech and Language Processing for Assistive Technologies (SLPAT), held in Dresden, Germany, on 11 September, 2015. We received 35 paper submissions, of which 22 were chosen for oral presentation and one was a system demonstration; all 23 papers are included in this volume. This workshop was intended to bring researchers from all areas of speech and language technology with a common interest in making everyday life more accessible for people with physical, cognitive, sensory, emotional or developmental disabilities. This workshop builds on five previous such workshops (co-located with conferences such as ACL, NAACL, EMNLP and Interspeech); it provides an opportunity for individuals from research communities, and the individuals with whom they are working, to share research findings, and to discuss present and future challenges and the potential for collaboration and progress. While Augmentative and Alternative Communication (AAC) is a particularly apt application area for speech and natural language processing technologies, we purposefully made the scope of the workshop broad enough to include assistive technologies (AT) as a whole, even those falling outside of AAC. Thus we have aimed at broad inclusivity, which is also manifest in the diversity of our Program Committee. We are also very delighted to have Prof. Jonas Beskow from the Royal Institute of Technology, Stockholm, Sweden, as invited speaker. The success of this SLPAT 2015 edition was due to the authors who submitted such interesting and diverse work and which generated so intense discussions. Finally, we must thank all the people who made this event possible: The members of the Program Committee for completing their reviews promptly, and for providing useful feedback for deciding on the program and preparing the final versions of the papers. The Interspeech organisers who, in many ways, made the organisation easier, the ISCA Administrative Secretariat for handling finance and the Dresden University of Technology which hosted the event in their premises.}, author = {Alexandersson, Jan and Altinsoy, Ercan and Christensen, Heidi and Ljunglöf, Peter and Portet, François and Rudzicz, Frank}, year = {2015}, publisher = {Association for Computational Linguistics}, ISBN = {978-1-941643-79-2}, } @misc{alexandersson-etal-2013-proceedings-182221, title = {Proceedings of SLPAT-2013: 4th Workshop on Speech and Language Processing for Assistive Technologies}, abstract = {We are pleased to bring you the Proceedings of the Fourth Workshop on Speech and Language Processing for Assistive Technologies (SLPAT), held in Grenoble, France on the 21st and 22nd of August, 2013. We received 23 paper submissions, of which 12 were chosen for oral presentation and another 5 for poster presentation. In addition, two demo proposals were accepted. All 19 papers are included in this volume. This workshop was intended to bring researchers from all areas of speech and language technology with a common interest in making everyday life more accessible for people with physical, cognitive, sensory, emotional or developmental disabilities. This workshop builds on three previous such workshops (co-located with NAACL HLT 2010, EMNLP in 2011, and NAACL HLT 2012) and includes a special topic, "Speech Interaction Technology for Ambient Assisted Living in the Home", which is a follow-up of two events (ILADI 2012 co-located with JEP-TALN-RECITAL 2012 and a special session in EUSIPCO 2012). The workshop provides an opportunity for individuals from research communities, and the individuals with whom they are working, to share research findings, and to discuss present and future challenges and the potential for collaboration and progress. While Augmentative and Alternative Communication (AAC) is a particularly apt application area for speech and Natural Language Processing (NLP) technologies, we purposefully made the scope of the workshop broad enough to include assistive technologies (AT) as a whole, even those falling outside of AAC. While we encouraged work that validates methods with human experimental trials, we also accepted work on basic-level innovations and philosophy, inspired by AT/AAC related problems. Thus we have aimed at broad inclusivity, which is also manifest in the diversity of our Program Committee. We are very delighted to have Prof. Mark Hawley from the University of Sheffield as invited speaker. In addition we continue our tradition of a panel of AAC users, who will speak on their experiences and perspectives as users of AAC technology. Finally, this year we also have a tour of the DOMUS "smart home" of the Laboratoire d'Informatique de Grenoble. Because of the many submissions and program points, we have for the first time extended the workshop to two full days. We would like to thank all the people and institutions who contributed to the success of the SLPAT 2013 workshop: the authors, the members of the program committee, the member of the organising committee and the invited speaker Mark Hawley. Finally, we would like to thank the Universities of Grenoble for sponsoring and hosting the workshop in the Laboratoire d'Informatique de Grenoble premises. }, author = {Alexandersson, Jan and Ljunglöf, Peter and McCoy, Kathleen and Portet, François and Roark, Brian and Rudzicz, Frank and Vacher, Michel}, year = {2013}, publisher = {Association for Computational Linguistics}, address = {Stroudsburg, PA}, ISBN = {978-1-937284-93-0}, } @misc{alexandersson-etal-2012-proceedings-182225, title = {Proceedings of SLPAT-2012: 3rd Workshop on Speech and Language Processing for Assistive Technologies}, abstract = {We are pleased to bring you the Proceedings of the Third Workshop on Speech and Language Processing for Assistive Technologies (SLPAT), held in Montreal, Canada on the 7th and 8th of June, 2012. We received 13 paper submissions, of which 8 were chosen for oral presentation and another 2 for demonstration presentation - all 10 papers are included in this volume. This workshop was intended to bring researchers from all areas of speech and language technology with a common interest in making everyday life more accessible for people with physical, cognitive, sensory, emotional or developmental disabilities. This workshop builds on two previous such workshops (co-located with NAACL HLT 2010 & EMNLP in 2011); it provides an opportunity for individuals from research communities, and the individuals with whom they are working, to share research findings, and to discuss present and future challenges and the potential for collaboration and progress. While Augmentative and Alternative Communication (AAC) is a particularly apt application area for speech and Natural Language Processing (NLP) technologies, we purposefully made the scope of the workshop broad enough to include assistive technologies (AT) as a whole, even those falling outside of AAC. While we encouraged work that validates methods with human experimental trials, we also accepted work on basic-level innovations and philosophy, inspired by AT/AAC related problems. Thus we have aimed at broad inclusivity, which is also manifest in the diversity of our Program Committee. We would also like to thank the members of the Program Committee for completing their reviews promptly, and for providing useful feedback for deciding on the program and preparing the final versions of the papers. Thanks also to the NACL organizers for guidance and support. Finally, thanks to the authors of the papers, for submitting such interesting and diverse work, and to the presenters of demos and commercial exhibitions. }, author = {Alexandersson, Jan and Ljunglöf, Peter and McCoy, Kathleen and Roark, Brian and Waller, Annalu}, year = {2012}, publisher = {Association for Computational Linguistics}, address = {Stroudsburg, PA}, ISBN = {978-1-937284-20-6}, } @misc{friedoken-etal-2011-proceedings-182227, title = {Proceedings of SLPAT-2011: 2nd Workshop on Speech and Language Processing for Assistive Technologies}, abstract = {We are pleased to bring you these Proceedings of the Second Workshop on Speech and Language Processing for Assistive Technologies (SLPAT), held in Edinburgh, Scotland on 30 July, 2011. We received 19 paper submissions, of which 9 were chosen for oral presentation and another 6 for poster presentation – all 15 papers are included in this volume. In addition, five demo proposals were accepted, and short abstracts of these demos are also included here. This workshop was intended to bring researchers from all areas of speech and language technology with a common interest in making everyday life more accessible for people with physical, cognitive, sensory, emotional or developmental disabilities. This workshop builds on the first such workshop (co-located with NAACL HLT 2010); it provides an opportunity for individuals from research communities, and the individuals with whom they are working, to share research findings, and to discuss present and future challenges and the potential for collaboration and progress. While Augmentative and Alternative Communication (AAC) is a particularly apt application area for speech and Natural Language Processing (NLP) technologies, we purposefully made the scope of the workshop broad enough to include assistive technologies (AT) as a whole, even those falling outside of AAC. While we encouraged work that validates methods with human experimental trials, we also accepted work on basic-level innovations and philosophy, inspired by AT/AAC related problems. Thus we have aimed at broad inclusivity, which is also manifest in the diversity of our Program Committee. We are very excited to have four invited speakers. Sylvia Grant, Darryal Stark and Greg McMurchie will speak on their experiences and perspectives as users of AAC technology. Norman Alm will chair this expert panel and facilitate discussion between the panel and workshop participants. Norman has a wealth of research experience in applying NLP technologies to AAC and we look forward to a truly interactive and informative session. We would like to thank all four speakers for taking the time to participate and provide their collective insight to the workshop. We would also like to thank the members of the Program Committee for completing their reviews promptly, and for providing useful feedback for deciding on the program and preparing the final versions of the papers. Thanks also to Marie Candito, Bonnie Webber and Miles Osborne for assistance with logistics and to Brian Roark for his guidance and support. Finally, thanks to the authors of the papers, for submitting such interesting and diverse work, and to the presenters of demos and commercial exhibitions. }, author = {Fried-Oken, Melanie and Ljunglöf, Peter and McCoy, Kathleen and Waller, Annalu}, year = {2011}, publisher = {Association for Computational Linguistics}, address = {Stroudsburg, PA}, ISBN = {978-1-937284-14-5}, } @techreport{bos-etal-2000-dialogue-99877, title = {Dialogue dynamics in restricted dialogue systems}, author = {Bos, Johan and Ericsson, Stina and Larsson, Staffan and Lewin, Ian and Ljunglöf, Peter and Matheson, Colin}, year = {2000}, } @inProceedings{burden-ljunglof-2005-parsing-46987, title = {Parsing linear context-free rewriting systems}, abstract = {We describe four different parsing algorithms for Linear Context-Free Rewriting Systems (Vijay-Shanker et al., 1987). The algorithms are described as deduction systems, and possible optimizations are discussed.}, booktitle = {IWPT'05, 9th International Workshop on Parsing Technologies}, author = {Burden, Håkan and Ljunglöf, Peter}, year = {2005}, } @techreport{milward-etal-2006-dynamic-47000, title = {Dynamic multimodal interface reconfiguration}, author = {Milward, David and Amores, Gabriel and Blaylock, Nate and Larsson, Staffan and Ljunglöf, Peter and Manchón, Pilar and Pérez, Guillermo}, year = {2006}, } @techreport{bringert-etal-2005-development-46993, title = {Development of multimodal and multilingual grammars: viability and motivation}, author = {Bringert, Björn and Cooper, Robin and Ljunglöf, Peter and Ranta, Aarne}, year = {2005}, } @inProceedings{ljunglof-2004-grammatical-10786, title = {Grammatical Framework and multiple context-free grammars}, abstract = {We show that there is a simple one-to-one correspondence between Grammatical Framework with context-free backbone and Multiple Context-Free Grammars (MCFG). Since the parsing complexity for MCFGs is known to be polynomial in the length of the input, we get the same result for context-free GF. }, booktitle = {FG-04, 9th Conference on Formal Grammar}, author = {Ljunglöf, Peter}, year = {2004}, } @techreport{ljunglof-2002-functional-10778, title = {Functional Programming and NLP}, abstract = {Most of today's NLP software is developed using either a logic programming language such as Prolog, or a low-level imperative language such as C or C++. In this paper I will try to argue why the paradigm of functional programming (as opposed to logic and imperative programming) matters for natural language processing.}, author = {Ljunglöf, Peter}, year = {2002}, publisher = {Chalmers University of Technology}, address = {Göteborg}, } @book{ljunglof-2002-pure-10783, title = {Pure Functional Parsing - an advanced tutorial}, abstract = {Parsing is the problem of deciding whether a sequence of tokens is recognized by a given grammar, and in that case returning the grammatical structure of the sequence. This thesis investigates di erent aspects of the parsing problem from the viewpoint of a functional programmer. It is conceptually divided into two parts, discussing the parsing problem from di erent perspectives; rst as a comprehensive survey of possible implementations of combinator parsers; and second as pure functional implementations of standard context-free parsing algorithms. The rst part of the thesis is a survey of the possible implementations of combinator parsers that have previously been suggested in the litterature, relating their dirrefent usages. A number of previously unknown parser implementations are also given, especially e cient for small and medium-sized natural language applications. The second part of the thesis de ne elegant and declarative, pure functional versions of some standard parsing algorithms for context-free grammars. The goal has been to implement the algorithms in a way that is close to their intuitive formulations, not sacri cing computational e ciency. The implementations only use simple data structures not relying on a global updateable state, thus opening the way for nice functional implementations. Finally the thesis implements parser combinators that can collect the grammatical structure in the program, to be able to use any suitable parsing algorithm and not just recursive descent. However, this requires an mildly impure extension of the host language Haskell.}, author = {Ljunglöf, Peter}, year = {2002}, publisher = {Chalmers University of Technology}, address = {Göteborg}, } @inProceedings{ljunglof-2003-abstract-10785, title = {An abstract view of generalized LR parsing}, booktitle = {NWPT-03, 15th Nordic Workshop on Programming Theory}, author = {Ljunglöf, Peter}, year = {2003}, } @book{ljunglof-2004-expressivity-10794, title = {Expressivity and Complexity of the Grammatical Framework}, abstract = {This thesis investigates the expressive power and parsing complexity of the Grammatical Framework (GF), a formalism originally designed for displaying formal propositions and proofs in natural language. This is done by relating GF with two more well-known grammar formalisms; Generalized Context-Free Grammar (GCFG), best seen as a framework for describing various grammar formalisms; and Parallel Multiple Context-Free Grammar (PMCFG), an instance of GCFG. Since GF is a fairly new theory, some questions about expressivity and parsing complexity have until now not been answered; and these questions are the main focus of this thesis. The main result is that the important subclass context-free GF is equivalent to PMCFG, which has polynomial parsing complexity, and whose expressive power is fairly well known. Furthermore, we give a number of tabular parsing algorithms for PMCFG with polynomial complexity, by extending existing algorithms for context-free grammars. We suggest three possible extensions of GF/PMCFG, and discuss how the expressive power and parsing complexity are influenced. Finally, we discuss the parsing problem for unrestricted GF grammars, which is undecidable in general. We nevertheless describe a procedure for parsing grammars containing higher-order functions and dependent types.}, author = {Ljunglöf, Peter}, year = {2004}, publisher = {Chalmers University of Technology}, address = {Göteborg}, ISBN = {91-628-6331-2}, } @article{ljunglof-2004-functional-10780, title = {Functional chart parsing of context-free grammars}, abstract = {This paper implements a simple and elegant version of bottom-up Kilbury chart parsing (Kilbury, 1985; Wir´en, 1992). This is one of the many chart parsing variants, which are all based on the data structure of charts. The chart parsing process uses inference rules to add new edges to the chart, and parsing is complete when no further edges can be added. One novel aspect of this implementation is that it doesnt have to rely on a global state for the implementation of the chart. This makes the code clean, elegant and declarative, while still having the same space and time complexity as the standard imperative implementations.}, journal = {Journal of Functional Programming}, author = {Ljunglöf, Peter}, year = {2004}, volume = {14}, number = {6}, pages = {669--680}, } @article{claessen-ljunglof-2000-typed-99880, title = {Typed logical variables in Haskell}, abstract = {We describe how to embed a simple typed functional logic programming language in Haskell. The embedding is a natural extension of the Prolog embedding by Seres and Spivey. To get full static typing we need to use the Haskell extensions of quantified types and the ST-monad.}, journal = {Electronic Notes in Theoretical Computer Science}, author = {Claessen, Koen and Ljunglöf, Peter}, year = {2000}, volume = {41}, number = {1}, } @techreport{bos-etal-1999-survey-99870, title = {Survey of existing interactive systems}, author = {Bos, Johan and Larsson, Staffan and Ljunglöf, Peter and Lewin, Ian and Matheson, Colin and Milward, David}, year = {1999}, } @inProceedings{lange-ljunglof-2018-mulle-274014, title = {MULLE: A grammar-based Latin language learning tool to supplement the classroom setting}, abstract = {MULLE is a tool for language learning that focuses on teaching Latin as a foreign language. It is aimed for easy integration into the traditional classroom setting and syllabus, which makes it distinct from other language learning tools that provide standalone learning experience. It uses grammar-based lessons and embraces methods of gamification to improve the learner motivation. The main type of exercise provided by our application is to practice translation, but it is also possible to shift the focus to vocabulary or morphology training.}, booktitle = {NLPTEA 2018, the 5th Workshop on Natural Language Processing Techniques for Educational Applications, Melbourne, Australia, 19th July 2018}, author = {Lange, Herbert and Ljunglöf, Peter}, year = {2018}, publisher = {Association for Computational Linguistics}, address = {Melbourne, Australia}, } @inProceedings{ljunglof-kjellberg-2018-interactive-274247, title = {Interactive correction of speech recognition errors: implementation and evaluation for English and Swedish}, booktitle = {SLTC 2018, the 7th Swedish Language Technology Conference, Stockholm, 7-9th November 2018}, author = {Ljunglöf, Peter and Kjellberg, J. Magnus}, year = {2018}, } @inProceedings{lange-ljunglof-2018-putting-274013, title = {Putting Control into Language Learning}, abstract = {Controlled Natural Languages (CNLs) have many applications including document authoring, automatic reasoning on texts and reliable machine translation, but their application is not limited to these areas. We explore a new application area of CNLs, the use of CNLs in computer-assisted language learning. In this paper we present a a web application for language learning using CNLs as well as a detailed description of the properties of the family of CNLs it uses.}, booktitle = {CNL 2018, the 6th International Workshop on Controlled Natural Language, Maynooth, Co Kildare, 27-28th August 2018; published as volume 304 of Frontiers in Artificial Intelligence and Applications}, author = {Lange, Herbert and Ljunglöf, Peter}, year = {2018}, publisher = {IOS Press}, address = {Amsterdam}, ISBN = {978-1-61499-904-1}, } @inProceedings{ljunglof-olsson-2009-trik-99885, title = {TRIK: en talande och ritande robot för barn med kommunikativa funktionshinder}, booktitle = {Presentation vid 8:e Västsvenska Kommunikationskarnevalen, 1–2 juni 2009}, author = {Ljunglöf, Peter and Olsson, Maria}, year = {2009}, } @inProceedings{derbring-etal-2009-subtts-148340, title = {SubTTS: Light-weight automatic reading of subtitles}, abstract = {We present a simple tool that enables the computer to read subtitles of movies and TV shows aloud. The tool works by reading subtitle files, which can be freely downloaded or extracted from a DVD using existing tools, and read the text aloud through a speech synthesizer. The target audience are people who have trouble reading subtitles while watching a movie, for example people with visual impairments and people with reading difficulties, such as dyslexia. The application will be evaluated together with user from these groups to see if this could be an accepted solution to their need. }, booktitle = {Proceedings of the 17th Nordic Conference of Computational Linguistics NODALIDA 2009}, author = {Derbring, Sandra and Ljunglöf, Peter and Olsson, Maria}, year = {2009}, } @inProceedings{ljunglof-etal-2012-free-168968, title = {A free and open-source tool that reads movie subtitles aloud}, abstract = {We present a simple tool that enables the computer to read subtitles of movies and TV shows aloud. The tool extracts information from subtitle files, which can be freely downloaded from the Internet, and reads the text aloud through a speech synthesizer. There are three versions of the tool, one for Windows and Linux, another for Mac OS X, and the third is a browser-based HTML5 prototype. The tools are freely available and open-source. The target audience is people who have trouble reading subtitles while watching a movie, including elderly, people with visual impairments, people with reading difficulties and people who wants to learn a second language. The application is currently being evaluated together with user from these groups. }, booktitle = {SLPAT'12: 3rd Workshop on Speech and Language Processing for Assistive Technologies}, author = {Ljunglöf, Peter and Derbring, Sandra and Olsson, Maria}, year = {2012}, } @inProceedings{ljunglof-2000-formalizing-10773, title = {Formalizing the dialogue move engine}, abstract = {In this paper we present a calculus for reasoning mathematically about rule-based dialogue systems so called dialogue move engines developed in the TRINDI project. The calculus is similar to term rewriting systems and dynamic logic. It is defined using monads, which are used for describing programming languages, and in functional programming to capture computations with side-effects.}, booktitle = {Götalog Workshop on Semantics and Pragmatics of Dialogue}, author = {Ljunglöf, Peter}, year = {2000}, } @techreport{ljunglof-etal-2006-multimodal-46998, title = {Multimodal grammar library}, author = {Ljunglöf, Peter and Amores, Gabriel and Cooper, Robin and Hjelm, David and Manchón, Pilar and Pérez, Guillermo and Ranta, Aarne}, year = {2006}, } @inProceedings{ljunglof-2016-towards-245594, title = {Towards interactive correction of speech recognition errors}, abstract = {In this project we explore how to make quick fixes to simple texts using as few interactions as possible. There are several situations where this could be useful, such as when you are driving (and don't have access to a keyboard), if your device is too small for a proper keyboard (such as a mobile phone), or if you have a communicative disability (e.g., cerebral palsy, visual impairment, or something else). The main contribution of this work is about improving the online interaction for a user who wants to correct speech recognition errors. The actual error correction algorithm that we have used -- based on the Levenshtein edit distance -- has not been in focus, and can probably be improved substantially.}, booktitle = {SLTC'16, 6th Swedish Language Technology Conference}, author = {Ljunglöf, Peter}, year = {2016}, } @inProceedings{lundalv-etal-2014-language-205192, title = {AAC Language Resources in the Mainstream}, abstract = {In order to provide language support to individuals requiring augmentative and alternative communication (AAC) we need linguistic resources suitably organized and represented, e.g., sign language material, symbol and image libraries suitable for multiple cognitive levels, as well as textual support in many languages. So far, these resources have been developed and maintained in separate and uncoordinated efforts, either by commercial or non-profit actors, and targeting different specific groups and needs. In the long run, this is a non-inclusive, ineffective and expensive way of proceeding, leading to limited benefit for AAC users and stake-holders, as well as for potential wider application. In a number of related efforts, work is underway to link free symbol libraries (at present Blissymbols and ARASAAC), where applicable using the Concept Coding Framework (CCF) technology, to common state-of-the-art lexical resources and language technology. The aim is to gradually establish a foundation for inclusive AAC support based on, and included in, mainstream and openly available lexical and language resources for wide use in all sectors of society. Following up results in this area from the European AEGIS project, work is now proceeding in cooperation between DART (centre for AAC and AT), Språkbanken (the Swedish Language Bank) and Centre for language Technology (CLT), University of Gothenburg, Sweden, and others. Results from this work will be presented and demonstrated, and the implications will be discussed. These, and several other signs of the time, strongly indicate that this is way to go. }, booktitle = {ISAAC-2014 Conference for the International Society for Augmentative and Alternative Communication}, author = {Lundälv, Mats and Derbring, Sandra and Ljunglöf, Peter}, year = {2014}, } @inProceedings{ljunglof-2014-shrdlite-205190, title = {ShrdLite: Semantic Parsing Using a Handmade Grammar}, abstract = {This paper describes my approach for parsing robot commands, which was task 6 at SemEval 2014. My solution is to manually create a compact unification grammar. The grammar is highly ambiguous, and relies heavily on filtering the parse results by checking their consistency with the current world. The grammar is small, consisting of not more than 25 grammatical and 60 lexical rules. The parser uses simple error correction together with a straightforward iterative deepening search. Nevertheless, with these very basic algorithms, the system still managed to get 86.1% correctness on the evaluation data. Even more interesting is that by making the parser slightly more robust, the accuracy of the system rises to 93.5%, and by adding one single word to the lexicon, the accuracy is boosted to 98.0%.}, booktitle = {SemEval-2014 Task 6: Supervised Semantic Parsing of Robotic Spatial Commands}, author = {Ljunglöf, Peter}, year = {2014}, } @inProceedings{angelov-ljunglof-2014-fast-205189, title = {Fast Statistical Parsing with Parallel Multiple Context-Free Grammars}, abstract = {We present an algorithm for incremental statistical parsing with Parallel Multiple Context-Free Grammars (PMCFG). This is an extension of the algorithm by Angelov (2009) to which we added statistical ranking. We show that the new algorithm is several times faster than other statistical PMCFG parsing algorithms on real-sized grammars. At the same time the algorithm is more general since it supports non-binarized and non-linear grammars. We also show that if we make the search heuristics non-admissible, the parsing speed improves even further, at the risk of returning sub-optimal solutions.}, booktitle = {EACL'14, 14th Conference of the European Chapter of the Association for Computational Linguistics}, author = {Angelov, Krasimir and Ljunglöf, Peter}, year = {2014}, } @inProceedings{ljunglof-2013-type-189694, title = {Type-based Human-Computer Interaction}, booktitle = {TYPES 2013: Types for Proofs and Programs}, author = {Ljunglöf, Peter}, year = {2013}, } @inProceedings{burden-etal-2013-enabling-181043, title = {Enabling Interface Validation through Text Generation}, booktitle = {VALID 2013 The Fifth International Conference on Advances in System Testing and Validation Lifecycle}, author = {Burden, Håkan and Heldal, Rogardt and Ljunglöf, Peter}, year = {2013}, ISBN = {978-1-61208-307-0}, } @techreport{ericsson-etal-2006-software-65252, title = {Software illustrating a unified approach to multimodality and multilinguality in the in-home domain}, author = {Ericsson, Stina and Amores, Gabriel and Bringert, Björn and Burden, Håkan and Forslund, Ann-Charlotte and Hjelm, David and Jonson, Rebecca and Larsson, Staffan and Ljunglöf, Peter and Manchon, Pilar and Milward, David and Perez, Guillermo and Sandin, Mikael}, year = {2006}, } @techreport{ljunglof-etal-2005-talk-65264, title = {The TALK Grammar Library: an Integration of GF with TrindiKit}, author = {Ljunglöf, Peter and Bringert, Björn and Cooper, Robin and Forslund, Ann-Charlotte and Hjelm, David and Jonson, Rebecca and Ranta, Aarne}, year = {2005}, } @inProceedings{ljunglof-2005-polynomial-10789, title = {A polynomial time extension of parallel multiple context-free grammar}, abstract = {It is already known that parallel multiple context-free grammar (PMCFG) [1] is an instance of the equivalent formalisms simple literal movement grammar (sLMG) [2, 3] and range concatenation grammar (RCG) [4, 5]. In this paper we show that by adding the single operation of intersection, borrowed from conjunctive grammar [6], PMCFG becomes equivalent to sLMG and RCG. As a corollary we get that PMCFG with intersection describe exactly the class of languages recognizable in polynomial time.}, booktitle = {LACL-05, 5th Conference on Logical Aspects of Computational Linguistics}, author = {Ljunglöf, Peter}, year = {2005}, } @article{bringert-etal-2005-multimodal-8752, title = {Multimodal Dialogue System Grammars}, abstract = {We describe how multimodal grammars for dialogue systems can be written using the Grammatical Framework (GF) formalism. A proof-of-concept dialogue system constructed using these techniques is also presented. The software engineering problem of keeping grammars for different languages, modalities and systems (such as speech recognizers and parsers) in sync is reduced by the formal relationship between the abstract and concrete syntaxes, and by generating equivalent grammars from GF grammars.}, journal = {Proceedings of DIALOR'05, Ninth Workshop on the Semantics and Pragmatics of Dialogue}, author = {Bringert, Björn and Ljunglöf, Peter and Ranta, Aarne and Cooper, Robin}, year = {2005}, } @techreport{ljunglof-etal-2006-enhanced-47003, title = {Enhanced multimodal grammar library}, author = {Ljunglöf, Peter and Amores, Gabriel and Burden, Håkan and Manchón, Pilar and Pérez, Guillermo and Ranta, Aarne}, year = {2006}, } @inProceedings{burden-etal-2013-opportunities-181040, title = {Opportunities for Agile Documentation Using Natural Language Generation}, booktitle = {ICSEA 2013 International Conference on Software Engineering Advances}, author = {Burden, Håkan and Heldal, Rogardt and Ljunglöf, Peter}, year = {2013}, } @inProceedings{ljunglof-siverbo-2012-bilingual-168967, title = {A Bilingual Treebank for the FraCaS Test Suite}, abstract = {We have created an open-source bilingual treebank for 99% of the sentences in the FraCaS test suite. The treebank was built in conjunction with associated English and Swedish lexica written in the Grammatical Framework Resource Grammar. The original FraCaS sentences are English, and we have tested the multilinguality of the Resource Grammar by analysing the grammaticality and naturalness of the Swedish translations. 86% of the sentences are grammatically and semantically correct and sound natural. About 10% can probably be fixed by adding new lexical items or grammatical rules, and only a small amount are considered to be difficult to cure.}, booktitle = {SLTC-2012, 4th Swedish Language Technology Conference, Proceedings of the Conference}, author = {Ljunglöf, Peter and Siverbo, Magdalena}, year = {2012}, } @inProceedings{cooper-etal-1999-accomodating-169112, title = {Accomodating Information States in Dialogue}, booktitle = {3rd Tbilisi Symposium on Language, Logic and Computation}, author = {Cooper, Robin and Engdahl, Elisabet and Larsson, Staffan and Ljunglöf, Peter}, year = {1999}, } @article{larsson-etal-1999-information-99872, title = {Information states and dialogue move engines}, abstract = {We explore the notion of information state in relation to dialogue systems, and in particular to the part of a dialogue system we call the dialogue move engine. We use a framework for experimenting with information states and dialogue move engines, which is being implemented in the form of TRINDIKIT, a toolkit for building dialogue move engines and dialogue systems. We also show how an experimental dialogue system (GoDiS) currently being developed in Göteborg within the framework can be provided with rules to handle accommodation of questions and plans in dialogue.}, journal = {Linköping Electronic Articles in Computer and Information Science}, author = {Larsson, Staffan and Cooper, Robin and Engdahl, Elisabet and Ljunglöf, Peter}, year = {1999}, volume = {4}, number = {23}, } @techreport{ljunglof-1998-konstruktiv-10763, title = {Konstruktiv mängdteori}, abstract = {Denna uppsats är ett försök att beskriva hur man kan formulera den traditionella mängdteorin konstruktivt. En del ickekonstruktiva axiom måste bytas ut mot alternativa varianter. Den teori som uppkommer är Peter Aczels konstruktiva Zermelo-Fraenkel. Det konstruktiva i teorin rättfärdigas genom tolkning i Martin-Löfs typteori.}, author = {Ljunglöf, Peter}, year = {1998}, publisher = {University of Gothenburg}, address = {Göteborg}, } @inProceedings{ljunglof-etal-2009-trik-91892, title = {TRIK: A talking and drawing robot for children with communication disabilities}, abstract = {This paper describes an ongoing project where we develop and evaluate setup involving a communication board (for manual sign communication) and a drawing robot, which can communicate with each other via spoken language. The purpose is to help children with severe communication disabilities to learn language, language use and cooperation, in a playful and inspiring way. The communication board speaks and the robot is able to understand and talk back. This encourages the child to use the language and learn to cooperate to reach a common goal, which in this case is to get the robot to draw figures on a paper.}, booktitle = {Proceedings of the 17th Nordic Conference of Computational Linguistics NODALIDA 2009}, author = {Ljunglöf, Peter and Larsson, Staffan and Thunberg, Gunilla and Mühlenbock, Katarina}, year = {2009}, volume = {4}, } @inProceedings{ljunglof-etal-2011-lekbot-148335, title = {Lekbot: A talking and playing robot for children with disabilities}, abstract = {This paper describes an ongoing project where we develop and evaluate a setup involving a communication board and a toy robot, which can communicate with each other via synthesized speech. The purpose is to provide children with communicative disabilities with a toy that is fun and easy to use together with peers, with and without disabilities. When the child selects a symbol on the communication board, the board speaks and the robot responds. This encourages the child to use language and learn to cooperate to reach a common goal. Throughout the project, three children with cerebral palsy and their peers use the robot and provide feedback for further development. The multimodal interaction with the robot is video recorded and analyzed together with observational data in activity diaries. }, booktitle = {SLPAT'11: 2nd Workshop on Speech and Language Processing for Assistive Technologies}, author = {Ljunglöf, Peter and Berman, Alexander and Claesson, Britt and Ericsson, Stina and Kronlid, Fredrik and Müller, Ingrid Mattsson and Ottesjö, Cajsa}, year = {2011}, } @inProceedings{ljunglof-2011-editing-148337, title = {Editing Syntax Trees on the Surface}, abstract = {We describe a system for interactive modification of syntax trees by intuitive editing operations on the surface string. The system has a graphical interface, where the user can move, replace, add, and in other ways modify, words or phrases. During editing, the sentence is kept grammatical, by automatically rearranging words and changing inflection, if necessary. This is accomplished by combining constraints on syntax trees with a distance measure between trees.}, booktitle = {Nodalida'11: 18th Nordic Conference of Computational Linguistics}, author = {Ljunglöf, Peter}, year = {2011}, } @techreport{ljunglof-siverbo-2011-bilingual-168965, title = {A Bilingual Treebank for the FraCaS Test Suite}, abstract = {We have created a bilingual treebank for 99% of the sentences in the FraCaS test suite. The treebank is built together with an associated bilingual English-Swedish lexicon written in the Grammatical Framework Resource Grammar. The original FraCaS sentences are English, and we have tested the multilinguality of the Resource Grammar by analysing the grammaticality and naturalness of the Swedish translations. 86% of the sentences are grammatically and semantically correct and sound natural. About 10% can probably be fixed by adding new lexical items or grammatical rules, and only a small amount are considered to be difficult to cure. }, author = {Ljunglöf, Peter and Siverbo, Magdalena}, year = {2011}, } @inProceedings{ljunglof-2007-converting-47008, title = {Converting Grammatical Framework to Regulus}, abstract = {We present an algorithm for converting Grammatical Framework grammars (Ranta, 2004) into the Regulus unification-based framework (Rayner et al., 2006). The main purpose is to take advantage of the Regulus-to-Nuance compiler for generating optimized speech recognition grammars. But there is also a theoretical interest in knowing how similar the two grammar formalisms are. Since Grammatical Framework is more expressive than Regulus, the resulting Regulus grammars can be overgenerating. We therefore describe a subclass of Grammatical Framework for which the algorithm results in an equivalent Regulus grammar.}, booktitle = {SPEECHGRAM, Workshop on Grammar-Based Approaches to Spoken Language Processing}, author = {Ljunglöf, Peter}, year = {2007}, } @inProceedings{ljunglof-2012-practical-168969, title = {Practical Parsing of Parallel Multiple Context-Free Grammars}, abstract = {We discuss four previously published parsing algorithms for parallell multiple context-free grammar (PMCFG), and argue that they are similar to each other, and implement an Earley-style top-down algorithm. Starting from one of these algorithms, we derive three modifications – one bottom-up and two variants using a left corner filter. An evaluation shows that substantial improvements can be made by using the algorithm that performs best on a given grammar. The algorithms are implemented in Python and released under an open-source licence. }, booktitle = {TAG+11, 11th International Workshop on Tree Adjoining Grammar and Related Formalisms}, author = {Ljunglöf, Peter}, year = {2012}, } @incollection{ljunglof-wiren-2010-syntactic-99884, title = {Syntactic parsing}, abstract = {This chapter presents basic techniques for grammar-driven natural language parsing, that is, analysing a string of words (typically a sentence) to determine its structural description according to a formal grammar. Basic parsing concepts are explained after which a number of well-known parsing techniques are described.}, booktitle = {Handbook of Natural Language Processing, 2nd edition}, author = {Ljunglöf, Peter and Wirén, Mats}, year = {2010}, publisher = {CRC Press, Taylor and Francis}, ISBN = {978-1420085921}, } @techreport{larsson-etal-1999-trindikit-99871, title = {TrindiKit 1.0 manual}, author = {Larsson, Staffan and Bos, Johan and Ljunglöf, Peter and Traum, David}, year = {1999}, } @inProceedings{ljunglof-2011-grasp-148499, title = {GRASP: Grammatikbaerad språkinlärning}, booktitle = {Presentation at the ISAAC Nordic Research Conference, 24 May 2011, Gothenburg}, author = {Ljunglöf, Peter}, year = {2011}, } @inProceedings{ljunglof-2011-edit-148504, title = {How to edit syntax trees on the surface}, booktitle = {Presentation at the LOCI Workshop II: Type Dependency, Type Theory with Records, and Natural-Language Flexibility, 16-17 June 2011, London}, author = {Ljunglöf, Peter}, year = {2011}, } @inProceedings{ljunglof-larsson-2008-grammar-73729, title = {A grammar formalism for specifying ISU-based dialogue systems}, abstract = {We describe how to give a full specification of an ISU-based dialogue system as a grammar. For this we use Grammatical Framework (GF), which separates grammars into abstract and concrete syntax. All components necessary for a working GoDiS dialogue system are specified in the abstract syntax, while the linguistic details are defined in the concrete syntax. Since GF is a multilingual grammar formalism, it is straightforward to extend the dialogue system to several languages. Furthermore, the GF Resource Grammar Library can be used to write a single concrete instance covering 13 different languages. }, booktitle = {Lecture Notes in Computer Science}, author = {Ljunglöf, Peter and Larsson, Staffan}, year = {2008}, } @inProceedings{larsson-etal-2000-godis-99874, title = {GoDiS – an accommodating dialogue system}, abstract = {This paper accompanies a demo of the GoDiS system. Work on this system was reported at IJCAI-99 (Bohlin et al., 1999). GoDiS is a prototype dialogue system for information-seeking dialogue, capable of accommodating questions and tasks to enable the user to present information in any desired order, without explicitly naming the dialogue task. GoDiS is implemented using the TRINDIKIT software package, which enables implementation of these behaviours in a compact and natural way.}, booktitle = {ANLP-NAACL'00 Workshop on Conversational Systems}, author = {Larsson, Staffan and Ljunglöf, Peter and Cooper, Robin and Engdahl, Elisabet and Ericsson, Stina}, year = {2000}, } @inProceedings{ljunglof-etal-1998-bamse-99865, title = {Bamse-land: A virtual theatre with entertaining agents based on well-known characters}, abstract = {The virtual theatre Bamse-land is an entertaining virtual world, where characters taken from the Swedish comic book Bamse are implemented as autonomous agents. The agents interact with each other according to their different personalities, which were derived from the comic. Our main concern has been to make an entertaining and believable application, and by keeping the algorithms non-complicated show that there can be easy solutions to the problem of designing believable agent applications. Experiences with users show that entertaining agents based on well-known characters bring many advantages, including a simplified design process and higher user involvement.}, booktitle = {Extended Abstracts of CHI'98, Conference on Human Factors in Computing Systems }, author = {Ljunglöf, Peter and Nilsson, Victoria and Siverbo, Magdalena}, year = {1998}, } @inProceedings{larsson-ljunglof-1999-godis-99866, title = {GoDiS and the dialogue move engine toolkit}, booktitle = {ACL'99, 37th Annual Meeting of the Association for Computational Linguistics}, author = {Larsson, Staffan and Ljunglöf, Peter}, year = {1999}, } @inProceedings{cooper-etal-1999-information-99868, title = {Information states and dialogue move engines}, booktitle = {IJCAI-99 Workshop on Knowledge and Reasoning in Practical Dialogue Systems}, author = {Cooper, Robin and Engdahl, Elisabet and Larsson, Staffan and Ljunglöf, Peter}, year = {1999}, } @techreport{larsson-etal-2000-trindikit-99878, title = {TrindiKit 2.0 manual}, author = {Larsson, Staffan and Berman, Alexander and Bos, Johan and Grönqvist, Leif and Ljunglöf, Peter and Traum, David}, year = {2000}, } @inProceedings{ljunglof-2009-trik-99886, title = {TRIK: en talande och ritande robot för barn med kommunikativa funktionshinder}, booktitle = {Presentation på ID-dagarna, 7–9 oktober 2009, Stockholm}, author = {Ljunglöf, Peter}, year = {2009}, } @inProceedings{ljunglof-2008-grammar-73727, title = {A grammar formalism for specifying ISU-based dialogue systems}, abstract = {We describe how to give a full specification of an ISU-based dialogue system as a grammar. For this we use Grammatical Framework (GF), which separates grammars into abstract and concrete syntax. All components necessary for a complete GoDiS dialogue system are specified in the abstract syntax, while the linguistic details are defined in the concrete syntax. Since GF is a multilingual grammar formalism, it is straightforward to extend the dialogue system to several languages.}, booktitle = {LONDIAL’08, 12th Workshop on the Semantics and Pragmatics of Dialogue}, author = {Ljunglöf, Peter}, year = {2008}, } @inProceedings{ljunglof-etal-2008-trik-91894, title = {TRIK: A talking and drawing robot for children with communication disabilities}, abstract = {This paper describes an ongoing project where we develop and evaluate setup involving a communication board (e.g., for Blissymbolics or manual sign communication) and a drawing robot, which can communicate with each other via spoken language. The purpose is to help children with severe communication disabilities to learn language, language use and cooperation, in a playful and inspiring way. The communication board speaks and the robot is able to understand and talk back. This encourages the child to use the language and learn to cooperate to reach a common goal, which in this case is to get the robot to draw figures on a paper.}, booktitle = {NordiCHI-08 Workshop: Designing Robotic Artefacts With User- and Experience Centred perspectives}, author = {Ljunglöf, Peter and Larsson, Staffan and Mühlenbock, Katarina and Thunberg, Gunilla}, year = {2008}, } @inProceedings{ljunglof-2008-transition-99881, title = {Transition-based parsing}, abstract = {In this paper we introduce a general framework for transition-based parsing algorithms. Among the algorithms that can be described in this framework are deterministic and generalized LR-parsing (Tomita, 1986), incremental tabular parsing such as the Earley algorithm (Earley, 1970), and projective and non-projective depenency parsing (Nivre, 2008).}, booktitle = {SLTC 2008: Proceedings of the Swedish Language Technology Conference}, author = {Ljunglöf, Peter}, year = {2008}, } @inProceedings{ljunglof-2009-dialogue-95890, title = {Dialogue Management as Interactive Tree Building}, abstract = {We introduce a new dialogue model and a formalism for limited-domain dialogue systems, which works by interactively building dialogue trees. The model borrows its fundamental ideas from type theoretical grammars and Dynamic Syntax. The resulting dialogue theory is a simple and light-weight formalism, which is still capable of advanced dialogue behaviour.}, booktitle = {DiaHolmia'09, 13th Workshop on the Semantics and Pragmatics of Dialogue}, author = {Ljunglöf, Peter}, year = {2009}, } @inProceedings{ljunglof-2010-grasp-130137, title = {GRASP: Grammar-based Language Learning}, abstract = {We are developing a pedagogical tool to support language learning and training for children with communicative disabilities. The system has a graphical interface, where the user can move, replace, add, and in other ways modify, words or phrases. The system keeps the sentence grammatical, by automatically rearranging the words and changing inflection, if necessary. In this way we hope that the system stimulates the child to explore the possibilities of language.}, booktitle = {SLTC-2010, 3rd Swedish Language Technology Conference}, author = {Ljunglöf, Peter}, year = {2010}, } @inProceedings{ljunglof-2010-trik-130134, title = {TRIK: A Talking and Drawing Robot for Children with Communication Disabilities}, abstract = {In this project we have developed and evaluated a setup involving a touch-screen computer with a dynamic screen software, and a drawing robot, which can communicate with each other via spoken language. The purpose is to help children with severe communication disabilities to learn language, language use and cooperation, in a playful and inspiring way. The communication board speaks and the robot is able to understand and talk back. This encourages the child to use language and learn to cooperate to reach a common goal, which in this case is to get the robot to draw figures on a paper. The robot has been tested on three children, two with cerebral palsy and one with autism spectrum disorder. During this session we present the preliminary results.}, booktitle = {ISAAC-2010, 14th Biennial Conference for Augmentative and Alternative Communication}, author = {Ljunglöf, Peter}, year = {2010}, } @inProceedings{ljunglof-2009-trindikit-99883, title = {trindikit.py: An open-source Python library for developing ISU-based dialogue systems}, abstract = {TrindiKit is one of the main tools for developing ISU-based dialogue systems, but it is implemented in a non-free dialect of the programming language Prolog. Therefore we have translated the TrindiKit toolkit into an open-source Python package. We have tried to remain close to the original TrindiKit formulation, while making the most of Python classes and objects.}, booktitle = {IWSDS'09, 1st International Workshop on Spoken Dialogue Systems Technology}, author = {Ljunglöf, Peter}, year = {2009}, } @misc{ljunglof-etal-2021-selected-306645, title = {Selected contributions from the Eighth Swedish Language Technology Conference (SLTC-2020), 25-27 November 2020}, abstract = {Selected extended papers from the Eight Swedish Language Technology Conference (SLTC-2020) which was held between 25-27 November 2020 in Gothenburg and online.}, author = {Ljunglöf, Peter and Dobnik, Simon and Johansson, Richard}, year = {2021}, publisher = {Linköping University Electronic Press}, address = {Linköping, Sweden}, ISBN = {978-91-7929-031-3}, } @incollection{lange-ljunglof-2021-learning-305146, title = {Learning Domain-Specific Grammars from a Small Number of Examples}, abstract = {In this chapter we investigate the problem of grammar learning from a perspective that diverges from previous approaches. These prevailing approaches to learning grammars usually attempt to infer a grammar directly from example corpora without any additional information. This either requires a large training set or suffers from bad accuracy. We instead view learning grammars as a problem of grammar restriction or subgrammar extraction. We start from a large-scale grammar (called a resource grammar) and a small number of example sentences, and find a subgrammar that still covers all the examples. To accomplish this, we formulate the problem as a constraint satisfaction problem, and use a constraint solver to find the optimal grammar. We created experiments with English, Finnish, German, Swedish, and Spanish, which show that 10–20 examples are often sufficient to learn an interesting grammar for a specific application. We also present two extensions to this basic method: we include negative examples and allow rules to be merged. The resulting grammars can more precisely cover specific linguistic phenomena. Our method, together with the extensions, can be used to provide a grammar learning system for specific applications. This system is easy-to-use, human-centric, and can be used by non-syntacticians. Based on this grammar learning method, we can build applications for computer-assisted language learning and interlingual communication, which rely heavily on the knowledge of language and domain experts who often lack the competence to develop required grammars themselves.}, booktitle = {Natural Language Processing in Artificial Intelligence—NLPinAI 2020}, author = {Lange, Herbert and Ljunglöf, Peter}, year = {2021}, publisher = {Springer International Publishing}, ISBN = {978-3-030-63787-3}, }