Google Translate: Swahili… bado kidogo

by

In the last couple of days Google has added Swahili to the list of languages supported by its Translate service. On one hand I’m very happy to see this addition as I think it has the potential to be a big step forward for development in east Africa. However, from first impressions the service still has a long way to go.

One of the main problems for Google is that Swahili is an agglutinative language – meaning that it puts morphemes (grammatical parts of words) together to form longer words. So it can be difficult for a machine to know where the morphemes (parts of the word carrying meaning) begin and end.

Here are some very simple examples that I tried putting into Google:

Swahili Morphemes English Google Translate
kupika ku-pika to cook cooking
ninapika ni-na-pika I am cooking I cooked
nilipika ni-li-pika I cooked I cooked
nitapika ni-ta-pika I will cook I cooked
sijapika si-ja-pika I have not cooked I cooked
apike a-pik-e let him cook apike
umepika u-me-pika you have cooked has cooked
tutapika tu-ta-pika we will cook we cooked
watakapopika wa-taka-po-pika when they will cook will kakopika
mlipokuwa mnapika m-li-po-kuwa m-na-pika when you (pl) were cooking as they were cooked
ikipikwa nasi i-ki-pik(w)a na-si if it is cooked by us it be boiled us
bado kidogo bado ki-dogo not quite yet still little

To be fair, from what I’ve seen the translations of single words isn’t bad at all. Where it falls down is in the grammar – translating Swahili past, present, future and negative-perfect tenses all to English past!

Going the other way, here a few English examples I tried:

English Swahili Google Translate English back-translation
many people watu wengi watu wengi many people
many trees miti mingi miti mingi many trees
many elephants tembo wengi wengi tembo many elephants
many cars magari mengi wengi magari many cars
I am cooking ninapika I am kupikia “I am” to cook with
I cooked nilipika mimi kupikwa I to be cooked
To be fair, from what I’ve seen the translations of single words isn’t bad at all. Kwa kweli, kutokana na yale ambayo nimeyaona, utafsiri wa maneno ya pekee siyo mbaya Kuwa na haki, kutokana na yale I’ve amemwona zote maneno ya wimbo sio mbaya wakati wote. In truth, coming from what “I’ve” he has seen all
words of song not bad all the time

At this point it looks to be a decent dictionary (although with nothing like the depth of the excellent Kamusi Project), and actually does ok with set phrases. However once you get past the set phrases that it knows it seems unable to understand the relatively simple grammar and come up with a meaningful translation.

This is obviously a work in progress, as the “Contribute a better translation” option shows. It would be interesting to know whether Google takes these user contributed translations and tries to work out how the grammars and structures of the languages compare, or whether it simply remembers the set translation in case anyone enters the exact same phrase again. The first would be fascinating to investigate, whereas I fear the second would be like trying to empty the ocean with a teaspoon.

Advertisements

Tags: , , , ,

17 Responses to “Google Translate: Swahili… bado kidogo”

  1. MattW Says:

    Interesting. Little sad you made tables but nice to reap the benefit of! One minor thing though, is that ‘kupika’ can also be ‘cooking’ as well as ‘to cook’ so they are not necessarily wrong there!

    • Mark Says:

      What do you have against tables Mr Wisbey? Yeah maybe I was a bit harsh on kupika… but it does highlight again the immense difficulties of machine translation between languages that have very different structures – in order to know which it was the machine would have to understand the context of the word in the sentence.

  2. Andy S Says:

    Interesting post. A good test for any translation is to translate from one language into another, then back to the first, and see how closely it matches.

    And I like your tables, they make me feel all 2003. 🙂

    • Mark Says:

      Yeah – maybe I should have done the two-way translation thing – I’m sure there would have been some interesting results!

      Glad the tables are providing entertainment to everyone. I wondered why WordPress didn’t have a table option and I had to manually enter the html… my web design skills are obviously stuck in 2003…! 🙂

  3. swahili Says:

    interesting post, thanks

  4. Beatrice Says:

    Good to know that Swahili is attaining the same status as any other foreign language. The tables are unavoidable when teaching many concepts in Swahili. I agree that the context has to be considered because one word could have so many meanings; kupika is essentially to cook BUT when a possessive now like kwake/his, kwao/theirs comes after kupika(kupika kwake) the meaning changes to his cooking/their cooking.

  5. richard Says:

    Kaka, asante sana kwa kunifungua macho kuhusiana na tafsiri ya tovuti ya google kwenda katika lugha ya kiswahili! ni jambo la maendelea sana kama ulivyo sema hapo awali. Lakini ni changamoto kwetu sisi Watanzania kuhakisha kwamba matumizi ya kiswahili ktk tovuti ya google yanaendelea na kuboreshwa zaidi.

  6. Chris Says:

    I haven’t spoken any Swahili in almost ten years so when I learned about the Google facility, I thought I’d test my memory.

    It can only really be useful as a dictionary tool, because of the very reasons you outline.

    That said, it did get ‘Mungu ibariki Tanzania’ first time of asking…

  7. Jack Says:

    Great comparison. i wonder if Swahili is a good sample and represents the quality of the Google translation in general.

  8. Sectional Garage · Says:

    google translator is great but sometimes there are so many errors on the translation ~-‘

  9. Johnstone Says:

    This is a great idea and a good effort. I pray that you get on well with the ideas.
    Good work there.
    Johnstone

  10. oj cordell Says:

    baba yabngu alikuwa missionari kutoka australia . alikaa tanzania kupita miake thelathini huko dodoma . miye nilizaliwa mjii arusha . sasa na kaa poland

  11. Abdulrahman Awes Mombasa Says:

    Ama kwa kweli huduma hii ya google translation itawasaidia wanafunzi wengi wa lugha ya kiswahili lakini wawe na mwalimu pia kando wa kuweza kusahihisha makosa ya hapa na pale yaliotokamana na tafsiri za mashine.
    Nataraji mabingwa na watalaam wa lugha hii tamu, watalishughulikia suala hili la kujaza mapengo haya hadi itapokuwa tarjuma ni ya sawa na kutegemeka.
    Mimi ni mkazi wa Mombasa,Kenya.

  12. Rashid Says:

    Jamani mimi Najaribu Kuchangia Kidoogo hapa Ushauri Wangu Kuhusu Google Translate, Tujaribu Kuifanya Mipango Au Kitu Ili sote Tufaidike Na Kurahisisha zaidi Ili tuweze Kupata Translation Ambayo Ifahamikayo Kwa inchi Au Lola Inayotumia hii Lugha ya kiswahili Ili iweze Kuwasaidiya Wote, Kwa mfano huu ,Ni lazima wapatikane Watu Watatu Au Wanne Mmoja kutoka Kenya ,Tanzania Burundi Na zanzibar Kila mmoja Aweke Translation Yake Kwenye Saver Ya Google Kwasababu Lugha Ya kiswahili Inatafauti Kidogo Tunapoandika Au tunapotafuta Kitu Inawezekana Kupata Maana nyengine Au pengine Hajawahi kutumiya yeye Hilo neno kwahiyo HII google Translate Iwe Inachaguwa wenyewe Aotomatic kwa kutokana na zile za madola manne au zaidi ili iweze Kutowa Translate Mzuri . Au Humu kwenye mtandao wa Internet iwe Ifanywe Uweze kuchaguwa Wewe mwenyewe Kwa Lugha Unayotaka Ama kenya, Tanzania,Burundi Ruwanda Or Zanzibar.

  13. Shantienergy Says:

    Ahsante kwa kazi kabambe. Ingawa umeitafsiri “apike” kama ulivyo, huenda ukajikuta taabani wakisema maana yake ni “Let her cook” English back-translation sijaelewa maana yake haswa wala yatoka wapi. Nadhani labda ni ile kwa kiingereza kama ‘direct translation’.
    Mimi mwenyewe ninajikakamua kujaribu kuwezesha wageni katika tovuti yangu ninayoiita Wiki-Hii waweze kutafsiri nene/jambo lolote mle kusoma kiswahili au lugha nyingine. Nikijariwa, watatafsiri na asiye kubali atafsiri hadi kila neno litakubalika. Suluhu ya Kenya, Tanzania n.k nimeliomba mumu humu-Ahsante sana Rashid.
    Ukweli ni hata kwetu Kenya watu wa pwani wanaogopwa kwani lugha yao ya mazungumzo hata hatuielewi kwani ni ya kuvutia mno.

  14. danny whitehouse, search engine optimisation, blog seo, online strategy, online marketing Says:

    danny whitehouse, search engine optimisation, blog seo, online strategy, online marketing…

    […]Google Translate: Swahili… bado kidogo « Under the Baobab Tree[…]…

  15. tulkojumi Says:

    tulkojumi…

    […]Google Translate: Swahili… bado kidogo « Under the Baobab Tree[…]…

Comments are closed.


%d bloggers like this: