Typology emerges from simplicity in representations and learning
We derive well-understood and well-studied subregular classes of formal languages purely from the computational perspective of algorithmic learning problems. We parameterise the learning problem along dimensions of representation and inference strategy. Of special interest are those classes of langu...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Institute of Computer Science, Polish Academy of Sciences
2021-08-01
|
| Series: | Journal of Language Modelling |
| Subjects: | |
| Online Access: | https://jlm.ipipan.waw.pl/index.php/JLM/article/view/262 |
| _version_ | 1818283243741380608 |
|---|---|
| author | Dakotah Jay Lambert Jonathan Rawski Jeffrey Heinz |
| author_facet | Dakotah Jay Lambert Jonathan Rawski Jeffrey Heinz |
| author_sort | Dakotah Jay Lambert |
| collection | DOAJ |
| description | We derive well-understood and well-studied subregular classes of formal languages purely from the computational perspective of algorithmic learning problems. We parameterise the learning problem along dimensions of representation and inference strategy. Of special interest are those classes of languages whose learning algorithms are necessarily not prohibitively expensive in space and time, since learners are often exposed to adverse conditions and sparse data. Learned natural language patterns are expected to be most like the patterns in these classes, an expectation supported by previous typological and linguistic research in phonology. A second result is that the learning algorithms presented here are completely agnostic to choice of linguistic representation. In the case of the subregular classes, the results fall out from traditional model-theoretic treatments of words and strings. The same learning algorithms, however, can be applied to model-theoretic treatments of other linguistic representations such as syntactic trees or autosegmental graphs, which opens a useful direction for future research. |
| first_indexed | 2024-12-13T00:33:49Z |
| format | Article |
| id | doaj.art-ea011a024c4b484d89a2ebfcce81a572 |
| institution | Directory Open Access Journal |
| issn | 2299-856X 2299-8470 |
| language | English |
| last_indexed | 2024-12-13T00:33:49Z |
| publishDate | 2021-08-01 |
| publisher | Institute of Computer Science, Polish Academy of Sciences |
| record_format | Article |
| series | Journal of Language Modelling |
| spelling | doaj.art-ea011a024c4b484d89a2ebfcce81a5722022-12-22T00:05:15ZengInstitute of Computer Science, Polish Academy of SciencesJournal of Language Modelling2299-856X2299-84702021-08-0191Typology emerges from simplicity in representations and learningDakotah Jay Lambert0Jonathan Rawski1Jeffrey Heinz2Department of Linguistics Institute for Advanced Computational Science Stony Brook UniversityDepartment of Linguistics, San José State UniversityDepartment of Linguistics Institute for Advanced Computational Science Stony Brook UniversityWe derive well-understood and well-studied subregular classes of formal languages purely from the computational perspective of algorithmic learning problems. We parameterise the learning problem along dimensions of representation and inference strategy. Of special interest are those classes of languages whose learning algorithms are necessarily not prohibitively expensive in space and time, since learners are often exposed to adverse conditions and sparse data. Learned natural language patterns are expected to be most like the patterns in these classes, an expectation supported by previous typological and linguistic research in phonology. A second result is that the learning algorithms presented here are completely agnostic to choice of linguistic representation. In the case of the subregular classes, the results fall out from traditional model-theoretic treatments of words and strings. The same learning algorithms, however, can be applied to model-theoretic treatments of other linguistic representations such as syntactic trees or autosegmental graphs, which opens a useful direction for future research.https://jlm.ipipan.waw.pl/index.php/JLM/article/view/262model theorysubregularitygrammatical inferenceformal language theoryphonologylearning complexity |
| spellingShingle | Dakotah Jay Lambert Jonathan Rawski Jeffrey Heinz Typology emerges from simplicity in representations and learning Journal of Language Modelling model theory subregularity grammatical inference formal language theory phonology learning complexity |
| title | Typology emerges from simplicity in representations and learning |
| title_full | Typology emerges from simplicity in representations and learning |
| title_fullStr | Typology emerges from simplicity in representations and learning |
| title_full_unstemmed | Typology emerges from simplicity in representations and learning |
| title_short | Typology emerges from simplicity in representations and learning |
| title_sort | typology emerges from simplicity in representations and learning |
| topic | model theory subregularity grammatical inference formal language theory phonology learning complexity |
| url | https://jlm.ipipan.waw.pl/index.php/JLM/article/view/262 |
| work_keys_str_mv | AT dakotahjaylambert typologyemergesfromsimplicityinrepresentationsandlearning AT jonathanrawski typologyemergesfromsimplicityinrepresentationsandlearning AT jeffreyheinz typologyemergesfromsimplicityinrepresentationsandlearning |