Skip to content

Latest commit

 

History

History
80 lines (61 loc) · 4.83 KB

README.md

File metadata and controls

80 lines (61 loc) · 4.83 KB
Raw

IndoCollex

meme
Welcome to our repository! This repository hosts the data on "IndoCollex: A Testbed for Morphological Transformation of Indonesian Word Colloquialism" Research Paper published on ACL-IJCNLP 2021. We also provide the guideline on how we annotate the data.

Repository Structure and Description

├───data
|   ├───full.csv
│   ├───formal_to_informal
│   └───informal_to_formal
├───dict
└───guideline
  1. data/formal_to_informal : Data to train our formal to informal system in the paper (contains train, dev, test)
  2. data/informal_to_formal : Data to train our informal to formal system in the paper (contains train, dev, test)
  3. data/full.csv : Full data that will be divided into formal_to_informal and informal_to_formal data. Some labels are also excluded.
  4. guideline: PDF guideline how we annotate the data
  5. dict: You can find formal-informal phrase level Indonesian dictionary (i.e. kamus alay), in tsv format.

Indonesian Colloquial Transformation

We break down colloqual transformation into several categories as follow:

Category Description Example
Disemvowelling elimination of some or all
the vowels		 kemarin - kmrn
belum - blum
besok - bsk
bagaimana - bgmn
Affixation modification, addition
or removal of affixes		 menyanyikan - nyanyiin
mengabari - ngabarin
Shortening shortening of the original word. sudah - dah
internet - inet
halusinasi - halu
Space/dash removal Space and dash removal,
including collapsing repeated words.		 di rumah - dirumah

terima kasih - terimakasih
ibu-ibu - ibu2
Sound alter slight change both in sound
and/or spelling in the text		 pakai - pake
pahit - pait
aku - akuh
Acronym Syllabic and letter compounds of
one or more words akin to
acronyms, abbreviations, and portmanteau		 ibu hamil - bumil
budak cinta - bucin
anak baru gede - abg
Reverse Letter reversal, or colloquially
known as “Boso Walikan”		 malang - ngalam
bang - ngab
Loan words* borrowed words, often from
local language or English.		 bapak - bokap
Jargon* tagline, terms that have been
made into a popular term		 mana saya tahu - meneketehe
  • We exclude these from our model data, but you can find them in the end-to-end formal-informal dictionary.

Some of Indonesian colloquial words are constructed by applying multiple transformation sequences, for example:

teman-teman -> teman2 -> temen2
bagaimana -> gimana -> gmn

Paper Link and Citation

You can find our paper here: https://aclanthology.org/2021.findings-acl.280.pdf

If you use any of our work for your academic work, please cite:

@inproceedings{wibowo-etal-2021-indocollex,
    title = "{I}ndo{C}ollex: A Testbed for Morphological Transformation of {I}ndonesian Word Colloquialism",
    author = {Wibowo, Haryo Akbarianto  and Nityasya, Made Nindyatama  and Aky{\"u}rek, Afra Feyza  and Fitriany, Suci  and Aji, Alham Fikri  and Prasojo, Radityo Eko  and Wijaya, Derry Tanti},
    booktitle = "Findings of the Association for Computational Linguistics: ACL-IJCNLP 2021",
    month = aug,
    year = "2021",
    address = "Online",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2021.findings-acl.280",
    doi = "10.18653/v1/2021.findings-acl.280",
    pages = "3170--3183",
}

Team

  1. Haryo Akbarianto Wibowo @ Kata.ai
  2. Made Nindyatama Nityasya @ Kata.ai
  3. ‪Afra Feyza Akyürek @ Boston University
  4. Suci Fitriany @ Kata.ai
  5. Alham Fikri Aji @ Kata.ai
  6. Radityo Eko Prasojo @ Kata.ai & Universitas Indonesia
  7. Derry Tanti Wijaya @ Boston University