Cleanup Project using SEDE Query

Question

What’s this post about?

We are a group of people who are doing some cleanup work. Currently, we are using an SEDE Query that finds unanswered¹ questions with answers by the OP. From a cleanup perspective, these are "low hanging fruit", since there is a better than average degree of certainty that the answer effectively answers the question. If the answers are good, we can upvote them so that the question is recognised as answered by the system. Otherwise, we can take whatever action is appropriate to deal with the question.

¹ an unanswered question is one that has no accepted or upvoted questions

Our work on this query has been a collaborative effort and learning experience, and as our process has developed and gained momentum, we felt we should explain to the community what we have been working on. This post is intended both as that explanation and as an invitation to you to get involved in our project if you’d like to.

Why?

We are doing this for the overall glory of Ask Ubuntu, because we are proud of this place we are part of and want it to be as useful and as polished as possible. Here are some benefits:

If questions have been answered well, the system should acknowledge that. Upvotes help everyone to find helpful answers. Neglected answers should get the credit they deserve. In many cases, answers that contain good solutions are poorly worded or formatted and can be edited into shape with a little effort from us.

On the other hand, old, unanswered questions don’t help anyone. As Shog9 says:

Unanswered questions are a dead-end in search; they're worth keeping around for a while on the chance that someone will find and answer them, but beyond a certain point if no one is expressing any interest in them they're just noise.

By closing and/or deleting questions that are not useful, we improve the signal-to-noise-ratio of our site and the quality of search results, and make space for questions about still-relevant topics to be asked anew, with a better chance of them receiving attention and answers.

Is this about boosting our answered rate? No, not directly. This is about giving each question the attention it deserves. If our answered rate benefits, that’s an added bonus :)

How?

Important - if you want to get involved, please visit us in the Downboat, before going to Trello and taking any actions. You can always ping me @Zanna with questions - I am around. We really want to know if anyone is doing anything on the board, or with the query directly, so we can keep track and check it.

We have been uploading the questions we are working on to a Trello board. Each card on the board represents and links to a question found by our query. We label the cards with our actions and use the lists to ensure each person’s work is checked by at least one other person. When we consider the question has been given the right treatment, we archive the card.

The actions we take are, briefly:

• Upvote an answer
• Leave a comment for the OP suggesting they improve and/or accept their answer
• Flag answers as NAA, and, if appropriate, downvote the question²
• Vote to close the question
• Add an answer

² This causes the question to be deleted by the system, and is potentially a controversial aspect of what we are doing, since in many cases the post is deleted with the agreement of only two non-mods, not three as is the case with delete votes.

For a detailed explanation, see this guide to the Trello Board which I guess is our "training" material to be read by anyone joining the project.

Even if you don’t have the time or inclination to help out directly, you can still help us a lot by reviewing (in the usual way) posts we’ve voted on or flagged or simply with your moral support. We know not everyone wants to do this kind of thing, and that’s completely fine, but we hope you’ll agree we are doing something positive for Ask Ubuntu.

We would like to note that the people working on the board at the moment are also answering, editing and voting on as many new questions as we can.

Future plans

Depending on how well this project goes, we might consider expanding it to work on other areas, such as questions with no answers, abandoned questions and so on.

Update

Since some suggestions have been made that our work is making a significant contribution to review queues, I am editing our post to add what I said in response to a comment: we have been doing this work for over a month, and in that time (due, I think, to efforts of reviewers spurred by Zacharee1) the close votes review queue dropped below 30. Our work makes a low impact on queues, in my opinion. We are sending an average of 5-10 posts daily to queues (perhaps a little more presently (~24 October) because of the stage we are at). I heartily encourage anyone who prefers not to review old posts to take full advantage of the friendly skip button.

Answer 1

From what I can tell, this project improves the site but also requires a lot of thankless work, so I would like to thank the contributors for their well-meaning efforts. I do hope that contributors will still have time to answer and curate new questions.

Answer 2

I quickly skimmed the question and think it's a wonderful idea!

However, Zacharee1 is constantly pointing out when there are >100 CV's in need of review and if we dig up Q&A's from three years ago it's making the total worse. With 3 y/o Q&A the original authors are might no longer around to defend what they wrote 3 years ago, or might not even remember the situation at that time. As such I think we should limit old post vetting only when current Q&A close candidates total <100.

The reason being I am only allow to cast 20 votes a day, even if I'm willing to cast 50 or even 100 in an hour or two, it's just not permitted by the software.

Perhaps overrides can be given for this project such that close votes on posts > 1 year old aren't counted towards the 20 day maximum? Perhaps a special section called "Old Q&A" where there are no limits on close votes?

I feel a lot of pressure to bring the totals down and it gets frustrating there is a 20 vote limit. The frustration builds with old Q&A where relevance is lowered when no one reads them, although they should be cleaned up for professionalism sake.