Re: [xml-dev] Have you used AI, machine learning, and/or dataanalytics to automatically map/convert one XML vocabulary to another?
No.
I cannot think of a situation where it would be viable, actually. I suppose if you went the path of using AI to figure out some domain for an element in one schema based on the values (ie city), then did the same for the other, and matched them, that might work. And by comparing names by some process of exclusion. And by figuring out what notation (lexical type) is being used for each element.
But it seems to me that these are not the kinds of mappings that we so often see: slightly different chages in domain, for example.
Here is an example, from XBRL. In XBRL, they are really strong that you should never just have a number for financial data: you need to know what currency, and at what date/time the number applies to (for exchange rates), and what period, the precision, and even which accounting rules were used to measure that fact. I don't think that AI deep learning systems would be much use there: unless you are so ancient that you include Expert Systems (sets of rules captured from humans) as part of AI.
Regards
Rick
