CommonMorph: Participatory Morphological Documentation Platform

Aso Mahmudi; Sina Ahmadi; Kemal Kurniawan; Rico Sennrich; Eduard Hovy; Ekaterina Vylomova

cs.CLApr 6, 2026

CommonMorph: Participatory Morphological Documentation Platform

Authors: Aso Mahmudi, Sina Ahmadi, Kemal Kurniawan, Rico Sennrich, Eduard Hovy, Ekaterina Vylomova

Organizations: The University of Melbourne · University of Zurich

Abstract

Collecting and annotating morphological data present significant challenges, requiring linguistic expertise, methodological rigour, and substantial resources. These barriers are particularly acute for low-resource languages and varieties. To accelerate this process, we introduce CommonMorph, a comprehensive platform that streamlines morphological data collection development through a three-tiered approach: expert linguistic definition, contributor elicitation, and community validation. The platform minimises manual work by incorporating active learning, annotation suggestions, and tools to import and adapt materials from related languages. It accommodates diverse morphological systems, including fusional, agglutinative, and root-and-pattern morphologies. Its open-source design and UniMorph-compatible outputs ensure accessibility and interoperability with NLP tools. Our platform is accessible at https://common-morph.com, offering a replicable model for preserving linguistic diversity through collaborative technology.

Figures & tables

Figure 1: The CommonMorph platform workflow facilitates elaboration of morphological structures by a linguist and provides an interoperable ecosystem for contributors to validate and enrich labelled databases.

Panel of Figure 2 — Figure 2: Linguists define (a) paradigm structures, (b) additional structural layers such as agreement, and (c) lexical entries. These components allow the generation of morphological forms that can be validated during elicitation. Linguists also create elicitation prompts (d), taking inspiration from examples produced by the system.

Figure 3: Example of a Latin verb conjugation table annotated with CommonMorph terminology: (1) An inflection class and (2) a paradigm structure within that inflection class; (3) a reusable layer contains several (4) reusable morphemes and their (5) morphosyntactic features; (6) a lemma and its (7) gloss and (8–9) different stems .

Panel of Figure 4 — Figure 4: Screenshots from the speaker interface.

Suggestion Model	Hawrami (Tewêle)	Kurdish (Sine)	Farsi (Tehran)	Turkish (Standard)	Spanish (Spain)	Arabic (Standard)
Linguist patterns	12.33	1.39	0.86	31.74	0.82	6.62
Linguist patterns + morphophonolgy	5.78	0.74	0.85	2.97	0.82	2.24
NN trained on 100 samples	48.41	54.23	37.25	23.31	20.98	41.93
NN trained on 500 samples	15.13	29.12	23.63	11.83	7.18	8.45
NN trained on 1000 samples	7.88	27.37	54.02	3.33	5.91	6.40
NN trained on 2000 samples	6.46	14.36	24.22	4.68	1.14	2.72

Table 1: Character Error Rate (CER, in %) of the inflection model in generating suggestions for different languages, which reflects the amount of correction required by speakers.

	Linguists		Speakers
Metric	Mean	SD	Mean	SD
User Learning Time	17.3 min	2.5	8.7 min	3.2
Contributing Time	1.0 hr	2.5	4.1 hr	9.2
Pedagogical Score	3.7/5	1.5	4.7/5	0.5
Efficiency Score	4.3/5	1.1	4.7/5	0.5
Satisfaction Score	4.7/5	0.5	4.7/5	0.5

Table 2: User survey results for the CommonMorph platform. The table reports average user learning and contribution times, as well as mean scores (1 = low, 5 = high) for pedagogical value, efficiency, satisfaction, and ease of use. SD indicates standard deviation.

Appendix figures & tables1 asset

Supplementary material from the paper’s appendix.

Appendix

Placeholder	Replaces with
<LEMMA>	hênan
<STEM1>	hêna
<STEM2>	hên
<SAMPLE.1.LEMMA>	girtin
<SAMPLE.1.STEM.1>	girt
<SAMPLE.1.STEM.2>	gir

Appendix

Table 7

Low-Resource Language Processing

Low-resource language processing builds and evaluates language technologies for languages with limited data, tools, or digital representation.

See all 565 papers

MorfFlex: Handling Rich Morphology
Jun 23, 2026Jaroslava Hlaváčová, Marie Mikulová, Barbora Štěpánková +2Morphological Analysis
Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics, Malostranské náměstí 25, 118 00 Prague 1, Czech Republic
MORPHOGEN: A Multilingual Benchmark for Evaluating Gender-Aware Morphological Generation
Apr 20, 2026Mehul Agarwal, Aditya Aggarwal, Arnav Goel +2Multilingual Language Model Evaluation Text Generation
SBILab, Indraprastha Institute of Information Technology Delhi
L-ReLF: A Framework for Lexical Dataset Creation
Mar 31, 2026Anass Sedrati, Mounir Afifi, Reda BenkhadraLow-Resource Language Processing
Department of Computer Science, KTH, Royal Institute of Technology, Stockholm, Sweden · Wikimedia Morocco, Rabat, Morocco

Jun 23, 2026J/K move · Enter open · S save

MorfFlex: Handling Rich Morphology

Jaroslava Hlaváčová, Marie Mikulová, Barbora Štěpánková, Milan Straka+1

Charles University, Faculty of Mathematics and Physics, Institute of Formal and Applied Linguistics, Malostranské náměstí 25, 118 00 Prague 1, Czech Republic

Morphological Analysis

Open paper AskPDF arXiv

We present MorfFlex, a morphological dictionary architecture suitable for languages with extensive regularity in both inflection and derivation. As the primary example of MorfFlex in use we introduce MorfFlex CZ, a morphological dictionary of Czech. It is distributed as a simple, unstructured list of <wordform, lemma, tag> triplets, however, its manually maintained, unpublished source files and conversion scripts encode a sophisticated system of inflectional and derivational patterns. These patterns dramatically reduce the otherwise enormous size of the dictionary, which currently contains over 100 million wordforms and more than 1 million lemmas. The MorfFlex CZ dictionary serves as an essential resource for ensuring the consistency of manual morphological annotation in the Prague Dependency Treebanks and underpins state-of-the-art automatic tools such as MorphoDiTa. In this paper, we focus on: (i) presenting an effective method for managing the rich morphological system within the dictionary, and (ii) demonstrating the utility of such a language resource for maintaining annotation consistency in corpora and supporting the development of advanced NLP applications.

Abstract

Low-Resource Language Processing

Explore similar work

MorfFlex: Handling Rich Morphology

MORPHOGEN: A Multilingual Benchmark for Evaluating Gender-Aware Morphological Generation

L-ReLF: A Framework for Lexical Dataset Creation