Well, if you understand under contextual semantic search the type of intelligent filtering of content described here https://towardsdatascience.com/automated-survey-processing-using-contextual-semantic-search-2dfe68338ab5, it definitely makes sense. Anyway I wouldn't call it semantic search it is just a kind of filtering. I have worked on some similar intelligent filtering on job offers using semantic web technology. No doubt, that it gives better results, than just syntactic filtering. However, be aware of its restrictions in contrast to human intellectual work.

From my perspektive there are two basic approaches to such a filtering mechanism. The first one is the traditional symbolic approach using synonyms, related terms, narrower terms etc. for identifying matching and non-matching content. For this approach you have to collect the relevant terms which identify the context of your interests and need to be aware that they probably are never complete. But it gives you sharp decisison boundaries.

The other class of approaches uses unsupervised word embedding techniques. They actually take the context of the words into account. For these approaches you don't have to describe the meaning of the terms sought or to collect a number of related terms, but you lose the sharp destinction boundary of symbolic approaches, which results in either to few or results not really related, but close to the content you like to see. Drawback is that they get weaker and weaker results the more meanings a term has.

I would consider these tools as what they are: just tools for supporting intellectual analysis, thus helping human analysist and relieving them from useless inspections.

µ:

Gratulations,,,:

You "dared to ask" [:-),)] what I call a SAQ

:= SAQ Seldom-Asked-Question

++++

#Before beeing able to answer this question,

it would be nice clarifiy some wordings for me.

# ...

#using a pseudo-"Turtle" Notation # from the Semantic-Web.

@prefix _

@question ?

:these-algorithms

_:do-you-have-a (link || reference)

.

_:time-constraints

_are-you-talking:about ( IT:performance bool:or [ IT:scalability-on "HPC-Systems"] )

.

Can you please go more into details on the following ?

effectiveness ##.. of-data:collected

_:what-do-you-have _:in-mind

#effectiveness (?perfomance, quality of the result, or soemthig else...)

.

in-comparison-to #qualitative/quantitative data.

_:what-do-you-have _:in-mind

# comparing which aspect (see above)

.

+++

Yours,

Dipl.-Phys. Frank.Haferkorn, OGIS OatGrain-InnovationS