deshanksuman's picture
Update README.md
7f6c6ce verified
metadata
language:
  - si
metrics:
  - accuracy
  - bleu
tags:
  - Romanized Sinhala
  - Sinhala
  - Transliteration
datasets:
  - deshanksuman/Swabhasha_RomanizedSinhala_Dataset
license: mit

Model Card for Model ID

This Repo is about Romanized Sinhala to Sinhala Transliteration using the Trie tree. This is work level transliteration and suggest you all the possible output.

This is best for use in Word Sense Disambiguation with Transliteration.

Model Details

print(translate("adrya",loaded_t))

Expected Output : ['ආදරය', 'ආදාරය', 'ආදරේය', 'අදාරය', 'අදරය']

Model Description

This model is capable of handling short-hand typing(Adhoc Transliteration).

eg Input: khmda Output : කොහොමද

If you are using this work: Kindly cite : T. G. D. K. Sumanathilaka, R. Weerasinghe and Y. H. P. P. Priyadarshana, "Swa-Bhasha: Romanized Sinhala to Sinhala Reverse Transliteration using a Hybrid Approach," 2023 3rd International Conference on Advanced Research in Computing (ICARC), Belihuloya, Sri Lanka, 2023, pp. 136-141, doi: 10.1109/ICARC57651.2023.10145648. keywords: {Terminology;Social networking (online);Computational modeling;Knowledge based systems;Message services;Data structures;Data models;Romanized Sinhala;Transliteration;Tri-gram;Rule-based;Prediction;Suggestion},

The datasets used can be accessed Through : https://github.com/Sumanathilaka/Swa-Bhasha-Sinhala-Singlish-Dataset

  • Developed by: Deshan sumanathilaka
  • Model type: rie Structure
  • Language(s) (NLP): Python
  • License: IEEE

Model Sources [optional]

  • Paper [optional]: T. G. D. K. Sumanathilaka, R. Weerasinghe and Y. H. P. P. Priyadarshana, "Swa-Bhasha: Romanized Sinhala to Sinhala Reverse Transliteration using a Hybrid Approach," 2023 3rd International Conference on Advanced Research in Computing (ICARC), Belihuloya, Sri Lanka, 2023, pp. 136-141, doi: 10.1109/ICARC57651.2023.10145648. keywords: {Terminology;Social networking (online);Computational modeling;Knowledge based systems;Message services;Data structures;Data models;Romanized Sinhala;Transliteration;Tri-gram;Rule-based;Prediction;Suggestion},

  • Demo [optional]: https://youtu.be/w6kdIDzoov4

Uses

Romanized Sinhala to Sinhala Transliteration Suggestion Level WSD for Transliteration

How to Get Started with the Model

Download all the files from the repo. You can open suggestion.ipynb file Call the function translate with an input String in Romanized Sinhala.

Model Card Authors [optional]

Deshan Sumanathilaka

Model Card Contact

deshankoshala@gmail.com