You know what? This is a good idea, this is a great idea. So great,
that I must make it in my inven-sys notes. Over time, when
maintaining my notes, especially my house logs, I strive to maintain
consistency. Unfortunately, keeping the language consistent isn’t
easy, being that as humans don’t have built-in knowledge of language,
it tends to drift overtime and increases the total world complexity,
even though the complexity for particular humans stays constant. So,
what is my proposed solution?
Well, if you ask me, keeping language and vocabulary consistent is something that is really only easy to do over at maximum a three year period. Information that must persist for longer than three years is going to need a way to compensate for language drift. That is, if you are writing formal computer languages in which the functionality will only work if the keywords match exactly.
So, here is my proposal, based off of long-term experience. Try to accumulate all synonyms for an identical concept. This has an added plus to it too. One problem with human natural language is that one word often times has more than one meaning assigned to it. So, what are you in effect doing when you attach synonyms? You are attaching multiple meanings to one concept. So, over time, what will happen is that even though each of those words may have one out of several meanings, in the long run, you’ll be able to use the general definitions of all of those words to cancel out the ambiguous meanings that are technically irrelevant and get to only a single meaning. This is a very interesting proposition indeed. And all you have to do is maintain a dictionary of synonyms. You don’t even need to worry about defining multiple meanings for one word. In fact, over time, you’ll be able to figure that out naturally by searching for all contexts that the word occurs in.
Why is this such a novel approach? It is a novel approach because, rather than relying on a complete and complex natural language processing framework, it works with only minimal computer knowledge, and best of all, it is easy to understand and explain to laypeople. Compare that with trying to explain the intricacies of the English language to laypeople. Although an average person’s brain is sophisticated enough to interpret the complexities of the English language directly and in real-time, if you ask people to start walking through a metalanguage that explains the English language for the sake of translating the rules into a format that corresponds to machine-executable code that can intelligently act on the English language, that gets way too complicated for people to understand, and thus, most people “give up” by, matter of fact, not even trying to help the computer science community build better software for interpreting natural language.
So, what’s the lesson to be learned here? Whenever you have an inexperienced person make queries against a computer system, the computer system will naturally prompt the user for words to disambiguate the query, some of which the user may not actually know. Although this creates a steeper learning curve for humans using the system, humans have the advantage of being able to learn new things quickly when they do not already know them in advance, especially in regard to language. Given that these terms come from human natural language, this also helps humans learn them faster, given that it is something familiar to what most people use on a daily basis.
Okay, great idea, but how is the database of synonyms built up to start with? The thing is, humans are the ones who change over time. You have to be able to have some humans to start with who change over time, but still remember the past, build up the relations. This is how things will normally be tied together.