Burmese Script Conversion utilizing Aksharamukha

This weblog submit is by Dr Adi Keinan-Schoonbaert, Digital Curator for Asian and African Collections, British Library. She’s on Twitter as @BL_AdiKS.

Interested in Myanmar (Burma)? Do you know that the British Library has a big assortment of Burmese materials, together with manuscripts courting again to the 17^th century, early printed books, newspapers, periodicals, in addition to present materials?

You possibly can search our primary on-line catalogue Explore the British Library for printed materials, or the Explore Archives and Manuscripts catalogue for manuscripts. However, to extend probabilities of discovering printed sources, you have to to go looking the Discover catalogue by typing within the transliteration of the Burmese title and/or writer utilizing the Library of Congress romanisation rules. Because of this trying to find an merchandise utilizing the unique Burmese script, or utilizing what you’d intuitively contemplate to be the romanised model of Burmese script, shouldn’t be going to get you very far (not but, anyway).

Excerpt from the Library of Congress romanisation scheme

The rationale for that is that that is how we catalogue Burmese assortment gadgets on the Library, following a coverage to transliterate Burmese utilizing the Library of Congress (LoC) guidelines. In concept, the advantage of this method particularly for Burmese is that it allows a two-way transliteration, i.e. the romanisation may very well be exactly reversed to present the Burmese script. Nevertheless, a significant situation arises from this romanisation system: romanised variations of Burmese script are to date faraway from their phonetic renderings, that almost all Burmese audio system are utterly unable to recognise any Burmese phrases.

With the LoC scheme being unintuitive for Burmese audio system, not reflecting the spoken language, British Library catalogue data for Burmese printed supplies find yourself nearly inaccessible to customers. And we’re not alone with this downside – different libraries worldwide holding Burmese collections and utilizing the LoC romanisation scheme, face the identical points.

The Buddha at Vesali in a Burmese manuscript, from the Henry Burney assortment. British Library, Or. 14298, f. 1

One helpful answer to this may very well be to search out or develop a device that converts the LoC romanisation output into Burmese script, and vice versa – just like how you’d use Google Translate. Maria Kekki, our Curator for Burmese collections, have found the net device Aksharamukha, which goals to facilitate conversion between varied scripts – additionally known as transliteration (transliteration into Roman alphabet is especially known as romanisation). It helps 120 scripts and 21 romanisation strategies, and by chance, Burmese is certainly one of them.

Aksharamukha: Script Converter screenshot

Utilizing Aksharamukha has already been of nice assist to Maria. As an alternative of painstakingly changing Burmese script manually into its romanised model, she may now copy-paste the conversion and make any needed changes. She additionally observed making fewer errors this manner! Nevertheless, it was lacking one essential factor – the flexibility to straight transliterate Burmese script particularly utilizing the LoC romanisation system.

Such performance wouldn’t solely save our curatorial and acquisitions workers a major period of time – but in addition assist some other libraries holding Burmese collections and following the LoC pointers. This might additionally permit Burmese audio system to search out materials within the library catalogue far more simply – readers may even use this platform to search out gadgets in our assortment, in addition to different collections all over the world.

To this finish, Maria received in contact with the developer of Aksharamukha, Vinodh Rajan – a pc scientist who can also be an skilled in writing programs, languages and digital humanities. Vinodh was glad to implement two issues: (1) add the LoC romanisation scheme as one of many transliteration choices, and (2) add areas in between phrases (in terms of spacing, in keeping with the LoC romanisation system, there are completely different guidelines for phrases of Pali and English origin, that are written collectively).

Vinodh demonstrating the brand new Aksharamukha performance, June 2022

Final month (July 2022) Vinodh carried out the brand new system, and what we will say, the result’s simply unbelievable! Readers at the moment are capable of copy-paste transliterated textual content into the Library’s catalogue search field, to see if we maintain gadgets of curiosity. It is usually a major enchancment for cataloguing and acquisition processes, having the ability to create acquisitions data and minimal data. As a subsequent step, we’ll look into updating all of our Burmese catalogue data to incorporate Burmese script (alongside transliteration), and contemplate the same plan of action for different South or Southeast Asian scripts.

I ought to point out that as a bonus, Aksharamukha’s codebase is absolutely open supply, is obtainable on GitHub and is properly documented. In case you have suggestions or discover any bugs, please be happy to boost a problem on GitHub. Thanks, Vinodh, for making this occur!