UD_Marathi-CMUPAN
Treebank card from UD release 2.18: metadata and texts come from UD_Marathi-CMUPAN/README.md; size is counted by streaming the .conllu files and checked against the release stats.xml. Generated by tools/treebanks build; do not edit by hand. Papers: Papers: UD_Marathi-CMUPAN; seeds: seeds/; converter: convert.md.
| Field | Value |
|---|---|
| UD name | UD_Marathi-CMUPAN |
| Language | Marathi (mr) |
| Family | IE |
| Treebank code | mr_cmupan |
| UD page | universaldependencies.org/treebanks/mr cmupan/index.html |
| License | CC BY-SA 4.0 |
| Genres | academic, news, grammar-examples, government |
| Text in data | yes |
| Parallel | no |
| In UD since | UD v2.18 |
| Authors | Pranav Kushare; Aditi Chaudhary; Luigi Talamo; Annemarie Verkerk; Helena Vaz |
| Contact | annemarie.verkerk@uni-saarland.de |
| Development | in the UD repo (here) |
Annotation layers
| Layer | Origin |
|---|---|
| Lemmas | automatic with corrections |
| UPOS | automatic with corrections |
| XPOS | automatic with corrections |
| Features | automatic with corrections |
| Relations | automatic with corrections |
UD scale: manual native — annotated by hand directly in UD; converted from manual — converted from manual annotation in another scheme; converted with corrections — converted and partly corrected by hand; automatic with corrections — automatic, with partial manual corrections; automatic — automatic; not available — the layer is absent.
Size
| Split | Sentences | Tokens | Words | Multiword tokens | Empty nodes |
|---|---|---|---|---|---|
| train | 7 616 | 94 144 | 94 144 | 0 | 0 |
| dev | 987 | 11 648 | 11 648 | 0 | 0 |
| test | 1 003 | 12 406 | 12 406 | 0 | 0 |
| total | 9 606 | 118 198 | 118 198 | 0 | 0 |
Tokens are surface tokens (Spanish vámonos is one), words are syntactic words (vamos + nos); multiword tokens are those split into several words; empty nodes are the N.M nodes of the enhanced graph.
Summary
This treebank is a modified version of a semi-automatically treebank authord by Aditi Chaudhary, which in turn is based on the treebanks released by KCIS, IIIT-Hyderabad. Additionally, the treebank also contains Marathi-Discourse: A manually annotated 35-sentence corpus covering political discourse.
Introduction
The sentences are based on treebanks released by KCIS, IIIT-Hyderabad . The primary goal of this version is to provide a clean, validated treebank by fixing structural inconsistencies, morphological feature errors, and dependency relation violations in the original treebank. The corrections were performed using python script-based semi-automatic logics. (For eg. Identifying and fixing invalid upos-deeprel, correcting lemmas, correcting/removing invalid morphological feature, etc). For transliteration Indic Transliteration python library was used.
Marathi-Discourse: The corpus also contains 35 Discourse sentences. Data for this is sourced from the official Marathi translation of Prime Minister (of India) Narendra Modi’s address to the nation regarding the COVID-19 pandemic, delivered on May 12, 2020.
To maintain traceability, each sentence ID in the carries a specific prefix:
cmupan_: Sentences from the CMU/IIIT-Hyderabad treebank.
DISC_: Discourse sentences.
This page in Markdown → https://md.mova.elamur.ai/free/en/docs/treebanks/mr/cmupan/