Re^4: [OT] How about an Off Topic Section?

What we need is a workable/working keyword system, so that rather than throwing things into a catch-all "off topic" bag, each post can have keywords that indicate the topic. New posts would maybe have the 'Perl' keyword selected by default. :-)

But lacking that, the stand-in is, and has always been, the free-text title of the post. It's quite effective. "Keyword searching" works very well indeed (if not 100% perfectly according to 100% of monks' wishes).

I reckon we are the only monastery ever to have a dungeon stuffed with 16,000 zombies.

Comment on Re^4: [OT] How about an Off Topic Section?

Replies are listed 'Best First'.
Re^5: [OT] How about an Off Topic Section? by ww (Archbishop) on Jun 05, 2015 at 21:34 UTC
I like the notion of a keyword system, but if the keywords are user-selected, why should we expect the results to be any better than our defacto (almost) keyword system of asking ~~perps...~~ uh, 'posters' of new root nodes to create clear, relevant and searchable titles? (Well, yes, I know we don't ask for the "searchable" but I think the frequency with which "haznav" crops up in a consideration for a title change ...seems to suggest that search-ability is a desideratum. and ...provides unambiguous support each of your second para's conclusions ...BUT at the same time suggests the extent to which users post weak, ambiguous or flat-out lousy titles is a reason for concern. But trying to write an AI process that would provide better titles is a challenge I'm not about to enlist for, and I suspect is too big a challenge and task for Monks who lack as huge excess of spare time. To that extent, a clearly labeled section for "Related topics" -- ( as suggested above ) -- seems to have merit, despite the number of times it's been rejected. Come, let us reason together: Spirit of the Monastery	[reply]

Replies are listed 'Best First'.

Re^5: [OT] How about an Off Topic Section?
by ww (Archbishop) on Jun 05, 2015 at 21:34 UTC

I like the notion of a keyword system, but if the keywords are user-selected, why should we expect the results to be any better than our defacto (almost) keyword system of asking ~~perps...~~ uh, 'posters' of new root nodes to create clear, relevant and searchable titles? (Well, yes, I know we don't ask for the "searchable" but I think the frequency with which "haznav" crops up in a consideration for a title change

...seems to suggest that search-ability is a desideratum.
and
...provides unambiguous support each of your second para's conclusions ...BUT at the same time suggests the extent to which users post weak, ambiguous or flat-out lousy titles is a reason for concern.

But trying to write an AI process that would provide better titles is a challenge I'm not about to enlist for, and I suspect is too big a challenge and task for Monks who lack as huge excess of spare time. To that extent, a clearly labeled section for "Related topics" -- ( as suggested above ) -- seems to have merit, despite the number of times it's been rejected.

Come, let us reason together: Spirit of the Monastery

[reply]