You are right, Rene, I split the thread.
Hmm.. I had hoped there would be a general understanding by now that Google translate should really not be used in this way. As Stephen Carlson says, it is really bad with Latin. Moreover, it will go to great lengths to make a coherent meaning even when there is none.
Google translate also leaves unrecognised words unchanged.
I also discovered a very annoying feature, of which I am not sure if it has been fixed by now.
If a text includes line breaks, it will translate each part between line breaks separately. Thus, a sentence may get broken up into potentially meaningless parts before being translated.
I'll see if I can find a way to test this.
1. the word 'totis' consists of the preceding word 'Taurus' 't and a As well as a possibility of an ending. In this case "-tis".
But "-lis, or -tus" could also be considered as an ending. So the possibility would also be " talis, tatus, etc. "
The true composition will be shown by the context.
2. I'm not claiming it's Latin.
The assignment "-um" ending would also work in German. Reichtum / Umgang / um das herum.
I can use it individually. End syllable, prefix and alone.
The variants with the characters "t,a,o,9", and the possible endings indicate a Latin possibility.
Personally, I consider pure Latin as we know it today unlikely. Nevertheless, it is romantically oriented.
3. I have read the works on the statistics of the last 20 years. They have done a good job, but they don't help me any more.
It's like being on a merry-go-round, but I prefer to go on a roller coaster.
4. experiments with comprehensible examples, and a constructive criticism seems to me to be the longer but better way.
I don't believe that the PC will ever bring a solution by probability calculations.
I know the Google translator is not the best.
I have been looking for a good translator for whole sentences for a long time.
A comparison only with single words is simply too uncertain.
[
attachment=4251]
It is important to place the individual characters under a rule in the system.
An arbitrary use would not be wise here.
In the example the arrangement is different, but it still makes sense.
So the "et and ut" have one thing in common in the sign.
In the same way, they can be placed individually or in words.
In VM there is still the "at and ot" application.
This spelling is used in German and Latin, so it is not new.
If I claim the most common character is "c" and the character "8" is a t from Taurus.
So I can write "c8" as "et" without violating the rules.
So the difference between hand A and hand B can be combined differently without breaking the rules and the same text is created.
Example / Change by arrangement
So it is necessary to produce a large amount of text in order to be able to start with a correction.
Nevertheless, it is better to work with a system from the outset. It is not important if it is the right system, as long as the rule works. Now I can start with corrections.
As R.Sale says "Trash in, trash out. But it is better to have something pre-sorted.
And there's the next problem.
If "89" = tum or tus.
So maybe it's sum or sus.
[attachment=4253]
Even if this statement is correct, I still do not know which "sus" to take.
Only for that I need a text.
1. get an overview. Bring order to chaos.
2. use good tools. PC programs, translators, and others.
3. determine the origin of the book, including the possible language. ( dialects )
4. interpret pictures correctly and create a possible text. Relationship between picture and word.
5. use the trash without inhibitions.
I'm pretty sure it's solvable. If several people were involved, they had to learn how to do it.
Why should she be more intelligent.
There, now I've loaded the memory enough.
