Stemming from a discussion here, there are many questions in less-frequented tags which are drive-by questions that never get an accepted answer, but are hugely useful and pop up quite often.
I came to stackoverflow because of a lack of resources to learn about google-charts and google-visualization around the web (there are newsgroup links, and here -- that's about it). Due to the lack of users in these particular tags, however, there is a lot of very useful information that hasn't been upvoted, and doesn't have any accepted answer. This is really tough for those of us who come here looking for help as it means we kind of have to wing it (try whatever is there and see if it works).
For instance, this question has been viewed over once per day over the past 9 months, has no upvotes, and no answer selected (it also happens to be one of the posts that helped me figure out how google-geochart works. To me, rather than having new users coming across these questions for the first time try to figure out which answers are useful, and which are not, having a community review of high-view unanswered/un-upvoted questions would greatly improve accessibility for more niche topics with fewer active users, and help build the quality of answers (and thereby the community).
There are so many places to go for more popular topics that catering to the niche topics will create a better resource for more people, and increase the usability/popularity of SO in the long term.
To create more accessible resources to folks who are finding these questions by providing guidance as to which answers are actually worth looking in to. This naturally happens for high-volume tags with many followers, but for less popular tags there is less natural review and fewer people able to review.
Questions that fit the following criteria will be flagged for review:
- At least 1 answer
- No upvotes on any answers
- No accepted answer
- Over X views (X would be decided looking at stats, or simply take the top Y% of views fitting the above criteria)
Review would result in one of the following actions:
- Determine that no answer actually solves the question (mark question as 'reviewed, no answer')
- Determine that one of the answers actually solves the question as described (mark answer as 'reviewed, best answer as selected by < reviewer name >')
- Determine that the question is outdated/no longer relevant due to changes in the topic at hand (mark question as 'potentially irrelevant due to < insert explanation here >')
Reviewers would be selected as:
- People with over X accepted answers for that tag
- People with over Y reputation
I'll let someone more clever figure out what reputation rewards (if any) should be doled out.
This will allow niche topics to have more comprehensive peer-reviewed questions/answers, improve the accessibility to newer users who can't judge which information is accurate/relevant, and improve the strength of the smaller communities by having good answers accepted in the long run through the review process.