6
$\begingroup$

A number of questions concern plotting/analysing large data sets. Therefore, it seems sensible to have a "large-data" tag. However, I don't like this name.

Any suggestions for a better name - or is "large-data" OK?

Update

I've started tagging questions with the "large-data" tag.

$\endgroup$
  • 1
    $\begingroup$ I kinda want to retag your post with a "tag-suggestion" new tag... $\endgroup$ – M. Tibbits Oct 8 '10 at 17:41
5
$\begingroup$

One important consideration is that the OP should easily be able to find appropriate tags when posting their question. The current interface is a text input box that performs a real-time prefix search of the tags. When someone is concerned about a large dataset, what might they start typing to find appropriate tags? Not "high-dimensional," I don't think! The prefix ought to be "large" or "big"--something simple like that. But that presumes the tag system uses prefixes as modifiers.

Alternatively, create a set of tags that modify the word "data" with suffixes. In this scheme, an OP concerned specifically about data might type the word "data" to produce a short list of data-related tags, such as "database", "database-large", "database-high-dimensional", "data-format", etc., allowing them to choose appropriately from this list.

$\endgroup$
  • $\begingroup$ I personally assume that they'll be prefixes - if I were trying to tag a question of this nature, "big-" would be my first attempt. $\endgroup$ – Matt Parker Oct 1 '10 at 21:07
  • $\begingroup$ I like the idea of suffixes, but it seems strange for some tags. For example, data-discrete, transformation-data, data-missing.. $\endgroup$ – csgillespie Oct 2 '10 at 20:43
  • $\begingroup$ I still haven't figured out how the synonyms work in practice. Would it be effective to make the suffix form of a tag a synonym of its prefix form? In particular, if we have a tag like discrete-data with a synonym data-discrete, then when a user types "data" in the tag search form, will it turn up "discrete-data"? $\endgroup$ – whuber♦ Oct 3 '10 at 13:16
  • $\begingroup$ Great idea! Use synonyms. To answer your question about how it works, play about with books and it's synonym textbook You can tag a question textbook and search for textbook and it will automatically get converted to books $\endgroup$ – csgillespie Oct 4 '10 at 9:59
5
$\begingroup$

I think "large-data" is good. I might have suggested "big-data", but that tends to mean explicitly data that cannot be kept in memory and/or must be dispersed (e.g. MapReduce).

$\endgroup$
  • $\begingroup$ I like big data, which I think extends to this question: e.g. see the wikipedia article: en.wikipedia.org/wiki/Big_data $\endgroup$ – ars Oct 1 '10 at 20:09
  • $\begingroup$ @ars That was my first inclination, but I tend to think of it as having a specific meaning very large data; maybe propose it as a separate answer so people can vote on it? $\endgroup$ – Shane Oct 1 '10 at 20:59
2
$\begingroup$

"big-data"

... for voting purposes. This is my preference, but it doesn't have the same bigger-than-memory connotation for me as it does for Shane, so best to put it to the collective mind...

$\endgroup$
0
$\begingroup$

What about just "big-n"? One could also envision related tags "big-p-small-n" & "big-n-small-p" -- not that the latter would be used all that much.

$\endgroup$
-1
$\begingroup$

"high-dimensional-data" ?

$\endgroup$
  • 2
    $\begingroup$ I tend to think of "large data" as referring to a big sample, and "high dimensional" as a reference to a large number of variables, irrespective of sample size. I could certainly be wrong, though. $\endgroup$ – Matt Parker Oct 1 '10 at 14:55
  • $\begingroup$ @Matt You're right, esp. for the precision "irrespective of sample size" (it applies to omics data, neuroimaging data, etc.). Then, I think it raises an interesting issue about scoping, and we should try to address both kind of configuration (huge sample or n<<p case) -- for the moment, I have no better idea :) $\endgroup$ – chl Oct 1 '10 at 15:02

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .