The first series of posts for the Omnivalence blog may be more technically inclined than a standard business blog offers. Data science already mystifies enough of those in management, so what could adding Quantum computing do to make things any clearer?
This series will to use introduce and develop Quantum data science (QDS) principles to explore Natural Language Processing (NLP), its current implementations, and explore some conceptual weaknesses regarding those implementations -- chiefly, language contact and the interaction of two languages with one another.
This need not be as abstract as two hypothetical spoken languages interacting at a large socio-historical scales -- one can consider a business considering a marketing campaign in (or expansion of operations into) a foreign market with a different language as a relevant-scale instance of language contact, where the three outcomes are primarily:
only one (possibly distinct language) is used to facilitate transactions;
a hybrid language emerges into a pidgin language;
the two languages are used interchangeably.
Now the reasons for why a firm may prefer their language to be adopted rather than accepted are beyond the scope of this example; this example is merely intended to illustrate that a business may wish to understand and anticipate which of these three outcomes may be most likely, and what actions/investments might influence that outcome, when making strategic decisions in human resource investments.
The motivating question for this blog series will be, what, if any advantage other than largely hypothetical algorithmic performance, would approaching this problem from QDS point of view add, with the primary difference between classical and quantum computing being that the former is in terms of Boolean logic gates, while the latter relies on superposition and the interaction of states in superposition. This motivating question is intended to focus the reader's (and blogger's) attention on whether some problems which may otherwise be intractable could be abstracted into tractable ones.
Comments