Report on the Progress in Bengali Language Testing
The Centre for Internet and Society (CIS) is doing a project on developing a text-to-speech software in 15 Indian languages. This blog post captures the key updates of the work done by CIS team for Bengali language.
Indian Bengali was added to eSpeak by the Indian team. The language was added and improved based on user testing and review.
- Conjunct consonant pronunciation when (ba) is conjunct was corrected.
- INHERENT vowel PRONUNCIEATION SOME TIMES NEED TO BE SUPPRESSED. It was made possible programmatically.
- Changed phoneme rules for য dZV modified to ্) যা :&
- sissue with য়/ Oj eg; desired pronunciation /Poi'sa but was being pronounced as 'payosa'
- Modified the [phoneme to jV and phone,me rules C)য় (C j
- Issue with 'ja'fala, that should double the pronunciation of the preceeding consonant, was rectified. Eg; Onnay was pronounced as 'onyai' as against the desired pronunciation
- The Bengali files sent by the developer on 20th December and Third January had been updated in the version.
- 71 Bengali Exception words had been included with their pronunciation rules - where inherent vowel even being between two consonant is not pronounced.
- Changed phoneme rules for য dZV modified to ্) যা :&
- issue with য়/ Oj was corrected. eg; desired pronunciation /Poi'sa but was being pronounced as 'payosa'
- Issue with 'ja'fala, that should double the pronunciation of the preceeding consonant was rectified. Eg; Onnay was pronounced as 'onyai' as against the desired pronunciation, The Bengali files sent by the developer on, 20th December and Third January had been updated in the version.
- 4/Auguest/2014Few exception words had been added দশটা , তিরিশটা, চল্লিশটা, পঞ্চাশটা, ষাটটা, সত্তরটা, করবে, বলবে, চলবে, শুনবে, শুনব, মানকচু, ওলকপি, ফটকিরি, চরণপদ্ম.
- In eSpeak Indian test build 1.11, the following exception words were added:
- অমরসঙ্গী O:mor'SongI
- চুপচাপ cup'c,ap
- পাগলা pa',g'la
- পাগলামো pa',g'la'mo
- মর্যাদা mO:rJ,ada
- আমপাতা 'ampa,ta
- আকাশবানী akaS'banI
- বিদঘুটে bi,dg#Ut.e
- দেবদেবী deb'debI
- অথচ O:t#oco
- আমরাই am,rai
- খেলা k#&'la
- খেলবো k#el'bo
- চারটি car,t.i
- একটি ekt.i
- একটা &kt.a:
- দরকার dO:r'kar'
- দেখছি dek#'c#i
- বলছি bo,lc#i
- করছি ko'rc#i
- শুনছি Sun'c#i
- ভাবছি b#a'bc#i
- ঘুরছি g#U,rc#i
- দিনরাত din'rat.
- সবরকম SO:b,rO:kom
- কাজকর্ম kaJ,kO:'rmo
- দুঃখ duk#'k#o
- তিনটে tin't.e
- চারটে car't.e
- পাঁচটা p~ac't.a
- সাতটা Sat't.a
- আটটা a:t.'t.a
- দশটা dOS't.:a
- একুশটা ekuS't.:a
- বাইশটা ba:iS't.:a
- তেইশটা tei'St.:a
- চব্বিশটা co'b'biS't.:a
- পঁচিশটা p~o'ciS't.:a
- ছাব্বিশটা ch:a'b'biS't.:a
- সাতাশটা Sat.aS't.:a
- আঠাশটা a:t.#a:S't.:a
- উনত্রিশটা un'triS't.:a
- তিরিশটা tiriS't.:a
- একত্রিশটা ek'triS't.:a
- বত্রিশটা bot:riS't.:a
- তেত্রিশটা tetriS't.:a
- চৌঁত্রিশটা c~ou:triS't.:a
- পঁয়ত্রিশটা p~oj'triS't.:a
- ছত্রিশটা chot:riS't.:a
- সাঁইত্রিশটা S~ai'triS't.:a
- আটত্রিশটা a:t.triS't.:a
- উনচল্লিশটা uno'col'l:iS't.:a
- চল্লিশটা col'l:iS't.:a
