Intermediate states between green and red

[geajack]

I had similar idea of this. I just checked if it would be good for me, but the redish page does not help me localize words that I don't know to practice them or pay more attention. Maybe it is a different purpose, but there is a website (https://bliubliu.com/en/) that the learning process was based on marking words as red when you don't know it. First, an AI marked in red according to a test made initially, then, it showed videos with transcript where you would mark as red or as green. Green words stop to appear as green to not pollute the text with colours, so that you focus more on still-learning-words and unknown-words. So, in case this project goes to other direction in a future... here are some ideas, not needing AI:

• I have my on texts here, so I could enter them as input and so these words would not appear as red (nor as green), would be just ignored, stay as in the original text;
• Add to the 'known words' all words of a selection;
• For words that are not in my vocabulary: mark as red;
• For words I translated and I am learning, mark as yellow; Maybe even include levels of difficulty for further ranking of words to practice by writing/speaking;
• If I see a red word or yellow that I feel confident, I click to mark as green (next texts this word will not have a color at all). I would like to participate of such a project, despite right now I am a bit busy trying to finish my phd and find a job. With the vocabulary builder, visualization tools could help too. Congratulations for the initiative! I saw your app been mentioned in the Canguro English video: https://www.youtube.com/watch?v=XbjGZHeLUn8

Originally posted by @heberleh in https://github.com/geajack/Wordology/issues/1#issuecomment-442949326

[geajack]

@heberleh I moved your suggestions here from the other thread. If I understand you correctly you're suggesting having intermediate states for words in between "unknown" and "known" which might be colored orange, yellow, etc. I thought about this during development. The problem is that unless you have some algorithm to update these states (which I doubt is feasible), then the user has to update all of those themselves, and I don't think many people (certainly not me) are going to want to have to keep going through all of their words and clicking on them to update those states manually.