The Voynich Ninja

I think I may have addressed this before. However I think it worth mentioning again given that Beata Megyesi and the people at the DECODE database, who I am often in contact with, are looking at potentially creating an automated transcription of the ciphers in their very large collection. I think this is the likely future of the transcription of the Voynich manuscript. As with most automated processes this would be much faster than could be done by a human being. I can foresee an advantage in this in that the user could potentially provide parameters to the transcription program which would affect the way it transcribed the text. For example one might specify that two glyphs are to be treated as the same or alternatively to be treated as different or for example certain glyphs are to be treated as a series of microglyphs or alternatively as one whole glyph. In short multiple different transcriptions could be produced each with different user defined preferences.If you don't like EVA then you could create your own transcription quickly and easily. However the program could potentially suggest which would be a better way to transcribe it e.g. suggest that these two glyph should be treated as one given that they verge into one another with no clear distinction. It could do this by analysing a vast number of instances of a glyph shape in a way that would be difficult for a human to do. I guess it could even some day potentially analyse the text to distinguish between the different authors.

Like many other areas of AI image recognition is progressing a pace and some day will catch up with the Voynich.

As others have already noticed I am quite interested in what the future of Voynich research will be like and how technological advances will continue to revolutionise it as they have in the past with new technologies like the internet.

I strongly believe there was a discussion on this (and of the associated difficulties) on VN, but right now I cannot point to the exact thread.

(28-02-2022, 04:48 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.I strongly believe there was a discussion on this (and of the associated difficulties) on VN, but right now I cannot point to the exact thread.

I think I remember that, but I think the difficulties are largely technological difficulties.

(28-02-2022, 05:46 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.I think I remember that, but I think the difficulties are largely technological difficulties.

Yes, exactly the technological difficulties, that's what I meant

(28-02-2022, 06:12 PM)Anton Wrote: You are not allowed to view links. Register or Login to view.

(28-02-2022, 05:46 PM)Mark Knowles Wrote: You are not allowed to view links. Register or Login to view.I think I remember that, but I think the difficulties are largely technological difficulties.

Yes, exactly the technological difficulties, that's what I meant

They will be resolved it is just a question of time. It is amazing how much areas of AI have advanced. Speech recognition and machine translation used to be very poor, now they seem quite impressive to me. Projects like those undertaken by Google DeepMind are very impressive. Of course there is a long way to go in AI. Nevertheless tasks like the automated transcription of the Voynich are quite "simple" compared to some of the more ambitious applications of AI.

There are numerous groups working on OCR of handwritten text. Some of these are also interested in the Voynich MS.
When the alphabet is not known beforehand (as with the Voynich MS) and the writing of consecutive lines partly overlaps, the problems are still insurmountable.
It will be a while, but it is certainly a worthwhile line of study.