Googletrans is a Python library that facilitates machine translation. Let’s see how it works.
Businesses are serving customers from all over the world these days and need to know how their products are viewed by them. This is why machine translation has become an important field of research in data science. It is crucial for global communication, breaking down language barriers in various domains. In education, it aids language learners and broadens access to diverse academic resources.
However, machine translation is one of the most complex tasks of natural language processing (NLP), mainly due to the subjective nature of different languages that may have hidden meanings and sarcasm. Google has researched this field extensively and built some great machine learning models that are state-of-the-art and great for our daily tasks.
Googletrans is an unlimited and free library in Python that Google has developed for the purpose of machine translation. It can detect languages as well as translate from one language to another very efficiently. It is compatible with all Python versions above 3.6, and is actively supported by Google with regular updates to improve its performance. Since it is free, it needs no configuration or authentication. It has a language detection feature as well, which is missing in other libraries. The output also has additional information like the score in confidence or the probable errors.
Now let us see how we can use googletrans to get some translations!
First, install this library using the following command on your terminal:
pip install googletrans==3.1.0a0
Once that this is done, let us look at the real translations. Open your favourite text editor and type the following code. Basically, I am asking googletrans to translate a line in Hindi into English.
from googletrans import Translator translator = Translator() translation=translator.translate(‘मेरा नाम िजु है’) print(translation.text) from googletrans import Translator translator = Translator() translation=translator.translate(‘मेरा नाम िजु है’) print(translation.text)
You can see below that the translation has been done:
(base) Jishnus-MacBook-Air: downloads jishnusaurav$ python translation.py my name is jishnu (base)Jishnus-MacBook-Air: downloads jishnusaurav$
If you remove the .text, you will get to see additional information like the source language, pronunciation, etc, as shown below:
(base) Jishnus-MacBook-Air: downloads jishnusaurav$ python translation.py
Translated(src=hi, dest-en, text-my name is jishnu, pronunciation=None, extra_data=”{‘translat...”)
If you want the destination language to be something else and not English, you can use the ‘dest’ argument, as shown below:
[(base) Jishnus-MacBook-Air: downloads jishnusaurav$ python translation.py
Translated (src=hi, dest-te, text=~pronunciation=Nā pēru jiṣṇu, extra_data=”{‘translat...”)
(base) Jishnus-MacBook-Air: downloads jishnusaurav$
Here I am asking it to translate the given line into an Indian language Telugu. It does the job and also gives us the pronunciation in Telugu. I can vouch for it that it is the right translation and pronunciation!!
You can also give the source language, if you know it, so that it identifies your script in that language only, using the ‘src’ argument.
Using this library, you can also do batch processing easily — just add the texts in the form of a list into the code, as shown below:
from googletrans import Translator translator = Translator() translation=translator.translate([‘मेरा नाम िजु’]) for a in translation: print(a.origin, ‘ -> ‘, a.text) print(translation)
The output, given below, shows that it gives us separate translations for each string:
[(base) Jishnus-MacBook-Air: downloads jishnusaurav$ python translation.py मेरा-> My नाम-> Name िजु-> Jishnu [<googletrans.models. Translated object at 0x1102c1350>, <googletrans.models. Translated object at 0x1102d1290>, <googletrans.models.Transl ated object at 0x1102dcb90>] (base) Jishnus-MacBook-Air: downloads jishnusaurav$
We can also use this library to detect languages by using the following code:
from googletrans import Translator translator = Translator() translation=translator.detect(‘मेरा’) print(translation)
In the output shown below, we can see that the language is detected and the confidence of how right that prediction is, is also shared with us.
(base) Jishnus-MacBook-Air: downloads jishnusaurav$ python translation.py Detected (lang=hi, confidence 0.98828125) (base) Jishnus-MacBook-Air: downloads jishnusaurav$
That’s it, you have learnt how to translate from one language to another using a Google library. You can now explore more libraries that perform machine translation, compare them and choose the best one for your use case. This article can act as a foundation for your journey into the NLP machine translation field!
