Lemmatisation of Word Additions

It would be nice to have the words automatically lemmatised (turning inflected words into their base form, e.g. “learnt” → “learn” or “shaking” → “shake”) so that learners can learn the base word rather than having to learn from the inflected word. This can be done by using an NLP library which usually has this as a feature.

This is mostly an issue also because users are currently unable to edit word additions, so learners who would prefer to have only the base words in their saved words are unable to edit the words to reflect that. It would be good to have this feature in conjunction with the feature suggested above also as a fallback in case the NLP lemmatiser gets it wrong.

@shouharuka Hey! What you described actually already works on the LR site. When you save a word like shook or said, it shows up under “Saved” as a basic form like shake or say.(In the screenshot: the way it was marked is the Context column, and the way it was saved is the Word column). So the lemmatization is already done. :slightly_smiling_face:

2 Likes

Hm, perhaps it only works with specific languages, because from Filipino (Tagalog) > English it doesn’t do this:

image
(the verb “iniisip” is the imperfective aspect conjugation of the verb “isipin”)

Maybe the NLP library or API that’s being used only supports some languages?

@shouharuka oh, okay, I should figure it out